DOUBLE-BRIDGE SWITCHING CONTROLLER

Description

Domaine technique

[0001] This description relates generally to electronic devices, in particular switching converters. Technique antérieure

[0002] Switching converters use one or more switches that alternately switch between on and off states at a switching frequency. They are used to provide voltage and / or current from a power supply with voltage / current values ​​different from the desired voltage / current values. For example, an AC-DC (alternating current-direct current) switching converter can obtain a direct current voltage from an alternating current source, such as a power grid or generator.

[0003] US patent 2021359609A1 discloses a method for controlling a converter comprising two H-bridges coupled by a transformer, in which: two switching sequences between states are respectively applied to the two bridges; one of the states of a first of the sequences corresponds to a given direction of application of a voltage to the transformer by the bridge to which said first sequence is applied; the first of the sequences varies during the same half-wave of an alternating voltage across the terminals of one of the bridges, such that: during at least a first period, the switching to and from said state occurs in the same state of a second of the sequences; and during at least a second period, the switching to and from said state occurs in different states of the second of the sequences. Summary of the invention

[0004] There is a need to overcome all or part of the disadvantages of known switching converters, and, in particular, of dual active bridge switching converters (DAB converters from the English "Dual Active Bridge Converter").

[0005] For example, there is a need for a control method for a switching converter, for example with double active bridges, allowing for zero voltage switching (ZVS), for example when the converter switches between two control modes.

[0006] One embodiment overcomes all or part of the disadvantages of known switching converters, and, in particular, of dual active bridge switching converters (DAB converters from the English "Dual Active Bridge Converter").

[0007] For example, one embodiment provides a method for controlling a switching converter, for example with double active bridges, allowing zero-voltage switching (ZVS) to be maintained, for example when the converter switches between two control modes.

[0008] One embodiment provides a method for controlling a converter comprising two H-bridges connected by a transformer, in which: repetitions, at a repetition frequency, of two switching sequences between several states are applied respectively to the two bridges; one of the states of a first of the two sequences corresponds to a given direction of application of a voltage to the transformer by the bridge to which the first of the two sequences is applied; in a first control mode, input and output switching of said one of the states of the first of the two sequences occurs at different states of the second of the two sequences; in a second control mode, input and output switching of said one of the states of the first of the two sequences occurs at the same state of a second of the two sequences;and each transition between the first mode at a nominal repetition frequency value and the second mode at the nominal repetition frequency value includes an adaptation of the repetition frequency determined by a first duration equal to a switching dead time duration or to the product of the switching dead time duration and a zero voltage switching margin.

[0009] According to one embodiment, each transition comprises, based on a model of the converter: the calculation of a first transferable power between the two bridges with the first mode and the nominal value of the repetition frequency, the first power being determined by the first duration; the calculation of a second transferable power between the two bridges with the second mode and the nominal value of the repetition frequency, the second power being determined by the first duration; the calculation of a third transferable power between the two bridges in the first and second modes at the nominal value of the repetition frequency and without taking into account the first duration; the calculation of a first value of the repetition frequency in the first mode, the first value of the repetition frequency being determined by the third transferable power and the first duration;and the calculation of a second value for the repetition frequency in the second mode, the second value of the repetition frequency being determined by the third transferable power and the first duration.

[0010] According to one embodiment: Each transition from the first mode to the second mode includes a jump from the first value to the second value of the repetition frequency; and each transition from the second mode to the first mode includes a jump from the second value to the first value of the repetition frequency.

[0011] According to one embodiment: each transition from the first mode with the nominal value of the repetition frequency to the second mode with the nominal value of the repetition frequency includes successively: in the first mode, switching from the first power to the third power by switching from the nominal value of the repetition frequency to the first value of the repetition frequency, in the third power, switching from the first mode to the second mode by switching from the first value of the repetition frequency to the second value of the repetition frequency, and in the second mode, switching from the third power to the second power by switching from the second value of the repetition frequency to the nominal value of the repetition frequency;and each transition from the second mode with the nominal value of the repetition frequency to the first mode with the nominal value of the repetition frequency comprises successively: in the second mode, switching from the second power to the third power by switching from the nominal value of the repetition frequency to the second value of the repetition frequency, in the third power, switching from the second mode to the first mode by switching from the second value of the repetition frequency to the first value of the repetition frequency, and in the first mode, switching from the third power to the first power by switching from the first value of the repetition frequency to the nominal value of the repetition frequency.

[0012] According to one embodiment: The two sequences each comprise two respective switching cycles of two branches of the bridge to which the sequence is applied; and the switching of the sequences takes place at times resulting from calculations based on an equality between: a setpoint power to be transferred between the bridges by the converter; and a transferable power calculated from the converter model and from values ​​of the voltages across the bridges.

[0013] According to one embodiment, the setpoint is calculated based on a voltage value received by one of the bridges and / or a voltage value to be supplied by the other of the bridges.

[0014] According to one embodiment, the setpoint is calculated so that the converter has a PFC type operation (from the English "Power Factor Correction").

[0015] According to one embodiment, said calculations are further based on a desired equality between values ​​of a current in the transformer at one of the switching instants of one of the two sequences and at one of the switching instants of the other of the two sequences.

[0016] According to one embodiment: The model includes a first parameter representing a second duration between switching instants of the two sequences in the first mode and a third duration between switching instants of the two sequences in the second mode; the first transferable power in the first mode is calculated from the model using a first value of the first parameter corresponding to an equality of the first and second durations; the third transferable power in the first and second modes is calculated from the model using a zero value of the first parameter; and the second transferable power in the second mode is calculated from the model using the first value of the first parameter corresponding to an equality of the first and third durations.

[0017] According to one embodiment: In the first mode, the model is determined by the following equation: P = V 1 2 4 . m . fsw . L − 4 − 4 . m 2 . x 2 + 4 . m 2 − 2 . m + 2 . x + m − m 2 where P is the transferable power, V1 is a voltage across one of the two bridges, x is the first parameter, fsw is the repetition frequency, L is a value of a leakage inductance of the converter, and m is a ratio between a ratio of the voltages across the two bridges and a turns ratio of the transformer; and in the second mode, the model is determined by the following equation: P = V 1 2 4 . fsw . L 1 − 2 . x 1 − m . 1 − 2 . x − 4 . x .

[0018] One embodiment provides a device configured to implement the above process.

[0019] One embodiment provides a converter comprising the above device. Brève description des dessins

[0020] These features and advantages, as well as others, will be described in detail in the following description of particular embodiments, given by way of non-limiting example, in relation to the attached figures, among which: there figure 1 represents, schematically, an example of a switching converter of a type to which the described embodiments apply; the figure 2 represents, schematically and in block form, an example of a converter control method for the figure 1 ; there figure 3 represents, schematically and in the form of timing diagrams, an example of an H-bridge switching sequence used in embodiments of converter control processes; the figure 4 represents, schematically and in the form of timing diagrams, an example of another H-bridge switching sequence used in embodiments of converter control processes; the figure 5 represents, schematically and in the form of timing diagrams, an example of how a control step of a switching converter can be implemented; and the figure 6 represents, schematically and in the form of timing diagrams, an example of how another control step of a switching converter can be implemented; the figure 7 represents, schematically and in block form, a method of implementing yet another step in controlling the converter of the figure 1 ; there figure 8 represents, schematically and by means of curves, a method of implementing yet another control step of the converter of the figure 1 ; and the figure 9 represents, schematically and in block form, one implementation method for the ordering step of the figure 8 . Description des modes de réalisation

[0021] The same elements have been designated by the same reference numerals in the different figures. In particular, structural and / or functional elements common to the different embodiments may have the same reference numerals and may have identical structural, dimensional and material properties.

[0022] For the sake of clarity, only the steps and elements necessary for understanding the described embodiments have been shown and are detailed. In particular, converter components such as switches, driver circuits, converter transformers, transformer leakage inductance, or capacitive elements are not described in detail, as the described embodiments are compatible with such components of a typical converter.

[0023] Unless otherwise specified, when referring to two connected elements, this means directly connected without any intermediate elements other than conductors, and when referring to two coupled elements, this means that these two elements can be connected or linked through one or more other elements.

[0024] In the description that follows, when referring to absolute positional qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative positional qualifiers, such as the terms "above", "below", "superior", "inferior", etc., or to orientational qualifiers, such as the terms "horizontal", "vertical", etc., unless otherwise specified, it refers to the orientation of the figures.

[0025] Unless otherwise specified, the expressions "approximately", "roughly", "about", and "on the order of" mean within 10%, preferably within 5%.

[0026] There figure 1 represents, schematically, an example of a 100-volt switching converter of a type to which the described embodiments apply. More specifically, the figure 1 is a reproduction of the figure 1 of patent FR 3 110 300, of patent application EP 3 910 775 and of patent application US 2021359609.

[0027] In the example shown, converter 100 receives a voltage V1 and provides a voltage V2.

[0028] The voltage V1 can be an alternating voltage, or AC voltage, for example, supplied by a source such as an electrical distribution network. The AC voltage can be sinusoidal or nearly sinusoidal. The AC voltage can have an RMS value of approximately 230 V or 110 V, and a frequency of 50 Hz or 60 Hz. The AC voltage can also come from an alternator. Alternatively, the voltage V1 can be a direct current voltage, for example, from a battery or photovoltaic cells.

[0029] The voltage V2 can be a direct current (DC) voltage, for example, from a battery. As an example, the converter 100 can then act as a battery charger from the voltage V1. The battery can be a vehicle battery, and the DC voltage typically varies between 250 V and 450 V during battery charging, but can reach higher values, for example, around 800 V. The converter is then designed to supply the battery with power typically between 1 kW and 30 kW during charging. Alternatively, the DC voltage can also be supplied to another, unshown, stage of the converter. Alternatively, the voltage V2 can be an alternating current (AC) voltage, for example, from a power grid (the converter then acting as an inverter) or from the windings of an electric motor.

[0030] The converter 100 includes an H-bridge 110, for example receiving the voltage V1. By H-bridge, we mean a parallel association of at least two branches (branches 111 and 112 for the H-bridge 110) between two terminals or nodes (nodes 114H and 114L for the H-bridge 110).

[0031] Each bridge branch is defined by a combination of two switches electrically connected in series between the branch terminals. In the H-bridge 110, branch 111 comprises switches T11H and T11L in series between nodes 114H and 114L, with switch T11H located on the side of node 114H. In the H-bridge 110, branch 112 comprises switches T12H and T12L in series between nodes 114H and 114L, with switch T12H located on the side of node 114H.

[0032] The converter 100 further includes an H-bridge 120, for example, supplying the voltage V2. The H-bridge 120 comprises two branches 121 and 122 electrically connected in parallel between terminals 124H and 124L. Branch 121 includes switches T21H and T21L in series between nodes 124H and 124L, with switch T21H located on the side of node 124H. Branch 122 includes switches T22H and T22L in series between nodes 124H and 124L, with switch T22H located on the side of node 124H.

[0033] In examples where the voltage V1 is an AC voltage, each of the switches T11H, T11L, T12H, and T12L is voltage bidirectional, meaning that in its off state, it is designed to prevent current from flowing in either direction across the switch. More generally, the switches in the H-bridge(s) that supply and / or receive an alternating voltage, between bridges 110 and 120, are voltage bidirectional. The described embodiments are compatible with common types of voltage bidirectional switches.

[0034] Each bidirectional voltage switch is controlled by applying a control signal with two levels, corresponding respectively to the switch's open and closed states. This application is not described in detail here, as the embodiments are compatible with the usual methods of applying such a control signal to a bidirectional switch.

[0035] Preferably, the switches T11H, T11L, T12H, T12L, T21H, T21L, T22H, and T22L of the two bridges are also current-bidirectional, meaning that each switch, in its conducting state, is adapted to allow current to flow in both directions. The described embodiments are compatible with common types of current-bidirectional switches.

[0036] The H-bridges 110 and 120 are connected by a transformer 130. In other words, the transformer has a winding 131 connecting two nodes 141 and 142 of one of the H-bridges (bridge 110) and another winding 132 connecting two nodes 151 and 152 of the other of the H-bridges (bridge 120). Nodes 141, 142, 151, 152 are series connection nodes of the switches, respectively T11H and T11L, T12H and T12L, T21H and T21L, T22H and T22L, of the respective branches 111, 112, 121 and 122 of the H-bridges. In the example shown, winding 131 has a phase point on the side of node 141 and winding 132 has a phase point on the side of node 152.

[0037] The transformer 130 includes a leakage inductance 135. In the example shown, the leakage inductance connects one terminal of the winding 131 to the node 141. The leakage inductance 135 may include one or more inductive elements, such as wound conductors, electrically in series with one and / or the other of the windings 131 and 132. The leakage inductance 135 may also, in whole or in part, result from incomplete magnetic coupling between the windings 131 and 132. In this case, the transformer 130 has, between its windings 131 and 132, a coupling coefficient strictly less than one.

[0038] The transformer has a turns ratio n (n:1) between windings 132 and 131. The turns ratio between a first winding and a second winding is understood to be the ratio between the number of turns in the second winding and the number of turns in the first winding. If the transformer 130 is disconnected from the device and a voltage is applied across winding 132, the turns ratio is typically approximately equal to the ratio between the voltage across winding 131 and the applied voltage. The turns ratio depends on the voltages involved and the power to be transferred by the converter. For example, the turns ratio n is between 0.5 and 2, for instance between 0.5 and 1, preferably on the order of 0.75.

[0039] In operation, in each branch, the two switches are controlled in reverse, that is, so that when one of the switches in the branch is in the conducting or closed state, the other switch in the branch is in the blocked or open state.

[0040] Preferably, in each of branches 111, 112, 121, and 122, the switches are alternately switched on and off at a switching frequency fsw. In other words, each branch is repeatedly switched alternately between a state where one of the branch's switches is on and another state where the other switch is on. Typically, at each switching, a dead time is provided, during which both switches of the switched branch are simultaneously off to prevent short-circuiting the branch terminals. Thus, both switches of a branch can be off simultaneously but are controlled not to be on at the same time.

[0041] As an example, when the voltage V1, respectively V2, is an AC voltage, the value of this voltage is considered constant over a switching period equal to 1 / fsw when the frequency of the voltage V1, respectively V2, is much higher, for example at least 10 times higher, preferably at least 100 times higher, than the switching frequency fsw.

[0042] In each bridge, the switching of the branches forms a switching sequence. The switching sequences of the two branches are repeated at the switching frequency. With each repetition of the switching sequences, the leakage inductance stores / releases energy, thus circulating this energy from one bridge to the other of the converter.

[0043] For this reason, the leakage inductor 135 has a variable voltage across its terminals, equivalent to a voltage V135 across the represented leakage inductor. More precisely, the voltage V135 is located between node 141 and winding 131. Calculating the value of the voltage V135 across the leakage inductor 135, according to the convention chosen to represent this leakage inductor, is within the grasp of a person skilled in the art, using the examples in this description.

[0044] As an example, the converter further includes a capacitive element 160 connecting terminals 114L and 114H of the H-bridge 110. The capacitive element 160 can consist of one capacitor and / or several capacitors in series and / or in parallel.

[0045] As an example, voltage V1 is applied between nodes 114H and 114L through an impedance 162.

[0046] As an example, the capacitive element 160 and the impedance 162 form a filter allowing to limit variations, at each switching, of the voltage V1 and / or the current supplied to the converter.

[0047] As an example, the converter further includes a capacitive element 170 connecting terminals 124L and 124H of the H-bridge 120. The capacitive element 170 can consist of one capacitor and / or several capacitors in series and / or in parallel.

[0048] As an example, the voltage V2 between nodes 124H and 124L is supplied by the converter through an impedance 172.

[0049] The capacitive element 170 and the impedance 172 form, for example, a filter allowing to limit variations, at each switching, of the voltage V2 and / or the current supplied by the converter.

[0050] The converter 100 also includes a control circuit 180 (CTRL). The control circuit 180 receives values ​​VV1 and VV2 representing the respective voltages V1 and V2. The values ​​VV1 and VV2 can be generated by any common device, not shown, for measuring a voltage between two terminals.

[0051] The control circuit 180 provides SIG control signals for switches T11H, T11L, T12H, T12L, T21H, T21L, T22H, and T22L. The control circuit 180 can be comprised of any device suitable for implementing a converter control method, and in particular for generating the SIG control signals.

[0052] For example, the control circuit 180 includes a digital data processing unit, such as a microprocessor, and a memory. The memory includes, for example, a program which, when read and executed by the processing unit, triggers the implementation of the converter control process, i.e., the generation of the SIG commands.

[0053] SIG commands are applied to switches in a usual way by circuits not shown, such as close control circuits and / or isolation circuits between reference potentials of the switch commands and the control circuit 180.

[0054] There figure 2 represents, schematically and in block form, an example of a 200 converter control method. figure 1 More specifically, the figure 2 is a reproduction of the figure 2 of the French patent and the aforementioned US and EP patent applications.

[0055] At step 201 (MEAS V1), the voltage V1 between terminals 114H and 114L is measured. This results in the value V V1 ( figure 1 ) representative of the voltage V1.

[0056] Alternatively, the voltage V1 is predefined. If the voltage V1 is alternating, step 201 can then consist of any step that generates a value representative of the values ​​of the voltage V1 as a function of time, for example, any step that generates values ​​evolving sinusoidally in phase with the voltage V1. The value V1 can also be a predefined constant, if the voltage V1 is direct and predefined.

[0057] At step 202 (MEAS V2), the voltage V2 between terminals 124H and 124L is measured. This results in the value V V2 ( figure 1 ) representative of the voltage V2. Steps 201 and 202 can be simultaneous.

[0058] At step 210 (P SET POINT), the control circuit 180 ( figure 1 ) receives the values ​​VV1 and VV2 obtained in steps 201 and 202. The control circuit 180 determines, i.e. calculates, from the values ​​VV1 and / or VV2, a power setpoint P* (not shown in figure 2 ), representative of the power to be transferred by the converter from the H 110 bridge to the H 120 bridge.

[0059] For example, the converter has a setpoint for the average power to be supplied over one or more half-cycles of the alternating voltage. This average power could correspond to the power to be supplied, for example, to a battery being charged. The setpoint P* can then be determined solely from the value V1.

[0060] This example is not limiting, and, in other examples that may correspond to a different use of a battery charge, the setpoint P* is determined from the values ​​V V1 and V V2.

[0061] For example, a determination of the power setpoint P* is described in the aforementioned French patent and US and EP patent applications, in relation to the figure 3 of these documents. In this example, the power setpoint P* is equal to a target power P1 that we want the converter to draw from the voltage source V1, less a power P160 representing the power supplied to the capacitive element 160 ( figure 1 ) during variations in voltage V1. The setpoint P* then corresponds to the power that we want the converter to draw from the voltage source V1, in order to obtain PFC (Power Factor Correction) operation of the converter, which means, for example, operation where the converter draws a sinusoidal current preferentially in phase with the voltage. PFC operation avoids creating a phase shift and / or harmonics in the current drawn relative to the input voltage.

[0062] The powers P1 and P160 are algebraic quantities that can each take positive and negative values. By algebraic quantity supplied to an element, we mean that when the algebraic quantity takes a positive value, this value is actually supplied to the element, and that when the algebraic quantity takes a negative value, this value is, in absolute terms, supplied by that element. The setpoint P* can also take positive and negative values. By negative setpoint of a power to be transferred from the bridge at H 110 ( figure 1 ) at the H 120 bridge ( figure 1 ), we mean that a power represented by the absolute value of this setpoint is to be transferred from the bridge in H 120 to the bridge in H 110. Put another way, we want energy to flow in both directions between the bridges of the converter.

[0063] After steps 201 and 202, at a step 220 (CALC ti), the control circuit 280 calculates instants ti, repeated at the switching frequency, of the switching to be applied to the switches T11H, T11L, T12H, T12L, T21H, T21L, T22H and T22L of branches 111, 112, 121 and 122.

[0064] To calculate the switching times ti, a converter model is used. Based on the voltages V1 and possibly V2, and the switching times ti, the model provides a prediction of the converter's operation. The model thus provides modeled, or predicted, values—that is, values ​​estimated from the model. These values ​​are, for example, currents, voltages, and / or powers in various converter components, such as switches or the transformer. The model is preferably designed so that, during operation, these currents, voltages, and / or powers take on values ​​that are substantially equal to, and preferably equal to, the modeled values.

[0065] In particular, the model provides a modeled value P of the power transferred by the converter from bridge 110 to bridge 120. The calculations performed in step 220 are such that the instants ti are those for which the modeled value allows the transfer of power equal to the setpoint P*, for example, such that the modeled value is equal to the setpoint value. In other words, the calculation of the instants ti is based on an equality between: the power value represented by the power setpoint P*; and a power calculated from the converter model 100 and the values ​​V V1 and possibly V V2.

[0066] At step 230 (APPLY SIG), the SIG control signals obtained in step 220 are applied to the converter switches. The SIG control signals ( figure 1 ) to be applied to the converter switches are provided by the control circuit 180 from the calculated instants ti.

[0067] The process of figure 2 is typically implemented repeatedly, with measurement steps 201 and 202 being carried out continuously, for example, and steps 210 and 220 being implemented in each loop of a program of device 180 ( figure 1 ).

[0068] There figure 3 represents, schematically and in the form of timing diagrams, an example of an SA sequence of H-bridge switching used in embodiments of control processes for a converter of the type of that of the figure 1 More specifically, the figure 3 is a reproduction of the figure 4 of the aforementioned patent and patent applications.

[0069] As an example, the SA switching sequence is applied to the bridge in H 110 ( figure 1 ). The SA switching sequence is repeated at the switching frequency.

[0070] The SA switching sequence comprises two switching cycles, SA1 and SA2, repeated in the respective branches 111 and 112 of the H-bridge 110. As an example, low (L) and high (H) levels are shown for each of the SA1 and SA2 switching cycles, corresponding to the on and off states, respectively, of switches T11H and T12H, in the branch to which the switching cycle is applied. In other words, the states shown for cycles SA1 and SA2 correspond to the commands of switches T11H and T12H, respectively. The commands of switches T11L and T12L, which are not shown, are, up to a delay, the inverse of the commands shown for switches T11H and T12H, respectively.

[0071] In other words, apart from the downtime, in the example where cycle SA is applied to bridge 110, for each of cycles SA1 and SA2: the high level corresponds to a conducting state of the respective switch T11H, T12H and to a blocked state of the respective switch T11L, T12L; and the low level corresponds to a blocked state of the respective switch T11H, T12H, and to a conducting state of the respective switch T11L, T12L.

[0072] Cycles SA1 and SA2 are preferably reversed. In other words, switches T11H and T12H of bridge 110, located on the same side of terminal 114H of bridge 110, are reversed. Similarly, switches T11L and T12L of bridge 110, located on the same side of terminal 114L of bridge 110, are reversed.

[0073] Thus, at a time tA1 of each repetition of the sequence SA, the sequence SA comprises two simultaneous switching of the cycles SA1 and SA2. At time tA1, within dead times, the bridge 110 switches, in other words flips, from a state N to a state P. In the state N, for two switches (T11H and T12H, or T12L and T11L) of the bridge 110 located on the side of the same terminal (respectively 114H or 114L) of the bridge 110, the blocked and conducting states are controlled respectively, and in the state P, the conducting and blocked states are controlled respectively for these two switches.

[0074] Similarly, at a time tA2 of each repetition of the sequence SA, bridge 110 switches from state P to state N with dead times nearby.

[0075] At each repetition of the switching sequence, the entry and exit times tA1 and tA2 of state P are located symmetrically with respect to an instant tAS. Alternatively, the instant tAS can be defined as the instant with respect to which the entry and exit times tA2 and tA1 of state N are located symmetrically.

[0076] Each of the cycles SA1 and SA2 has a duty cycle defined by the duration for which the cycle SA1, SA2 is at the control level of the respective switch T11H, T12H (high state). In the case of cycles SA1 and SA2 that are inverses of each other, the duty cycles of cycles SA1 and SA2, within the limits of any dead time, sum to 1.

[0077] Preferably, the duty cycles of SA1 and SA2 are approximately equal to 0.5, and more preferably equal to 0.5, within the limits of time intervals. In other words, the cycles SA1 and SA2, being inverses of each other, are also in opposite phase, within the limits of time intervals. It follows that the sequence SA has identical durations for both states N and P, within the limits of time intervals. At each repetition of the sequence SA, these identical durations are located symmetrically with respect to time tA1 or time tA2. Thus, the states N and P of the sequence SA are arranged symmetrically with respect to time tAS, within the limits of time intervals.

[0078] The SA switching sequence, described above in its application to the H 110 bridge, can be applied to the H 120 bridge ( figure 1 ) similarly, by replacing, with respect to the application to bridge 110 described above, branches 111 and 112 respectively with branches 121 and 122 ( figure 1 ). More specifically, for this, compared to the application to bridge 110 described above, switches T11H, T11L, T12H and T12L are replaced, respectively, by switches T21H, T21L, T22H and T22L.

[0079] There figure 4 represents, schematically and in the form of timing diagrams, an example of another H-bridge switching sequence used in embodiments of control methods for a converter of the type of that of the figure 1 More specifically, the figure 4 is a reproduction of the figure 5 of the aforementioned patent and patent applications.

[0080] As an example, the SB switching sequence is applied to the bridge in H 120 ( figure 1 ). The SB switching sequence is repeated at the same switching frequency as the SA sequence of the figure 4 .

[0081] The SB switching sequence comprises two switching cycles, SB1 and SB2, repeated in the respective branches 121 and 122 of the H-bridge 120. As an example, low (L) and high (H) levels are shown for each of the switching cycles SB1 and SB2, corresponding to the on and off states, respectively, of switches T21H and T22H in the branch to which the switching cycle is applied. In other words, the states shown for cycles SB1 and SB2 correspond to the commands of switches T21H and T22H. The commands of switches T21L and T22L, which are not shown, are, up to a delay, the inverse of the commands shown for switches T21H and T22H.

[0082] In other words, apart from the dead times, in the example where cycle SB is applied to bridge 120, for each of cycles SB1 and SB2: the high level corresponds to a passing state of the respective switch T21H, T22H and to a blocked state of the respective switch T21L, T22L; and the low level corresponds to a blocked state of the respective switch T21H, T22H, and to a blocked state of the respective switch T21L, T22L.

[0083] Each of the cycles SB1 and SB2 has a duty cycle defined by the duration for which the cycle SB1, SB2 is at the control level of the on state of the respective switch T21H, T22H (high level). Preferably, the duty cycles of cycles SB1 and SB2 are substantially equal to 0.5, more preferably equal to 0.5, within a few dead times.

[0084] The SB1 and SB2 cycles are preferably out of phase with each other. In other words, the SB1 and SB2 cycles have the same duty cycle and exhibit periods of 510 during which the SB1 and SB2 cycles are at different levels.

[0085] The result is that, with each repetition of the SB sequence: At time tB1, cycle SB2 switches from its high state to its low state. Bridge 120 switches from state N to state O; at time tB2, cycle SB1 switches from its low state to its high state. Bridge 120 switches from state O to state P; at time tB3, cycle SB2 switches from its low state to its high state. Bridge 120 switches from state P to state O; and at time tB4, cycle SB1 switches from its high state to its low state. Bridge 120 switches from state O to state N. The SB sequence then returns to its initial state before time tB1. The times tB1, tB2, tB3 and tB4 follow one another at each sequence, in this order or can be permuted according to the chosen initial state of the SB sequence.

[0086] The N and P states of bridge 120 correspond to the N and P states described in relation to the figure 3 That is to say, for the two bridge switches located on the same terminal of the bridge, in state N, the blocked and conducting states are respectively controlled, and, in state P, the conducting and blocked states are respectively controlled. State O corresponds to a state in which two bridge switches located on one terminal of bridge 120 are simultaneously in the conducting state, while the other two bridge switches (located on the other terminal of bridge 120) are in the blocked state. Thus, in state O, the configuration of the bridge switches corresponds to one of the following: the configuration in which switches T21H and T22H are conducting and switches T21L and T22L are blocked; and the configuration in which T21H and T22H are blocked and switches T21L and T22L are conducting.

[0087] At each repetition of the switching sequence, the entry and exit times tB2 and tB3 of state P are, apart from dead times, located symmetrically with respect to an instant tBS. Alternatively, the instant tBS can be defined as the instant with respect to which the entry and exit times tB4 and tB1 of state N are, apart from dead times, located symmetrically, or as the instant with respect to which the instants tB1 and tB2, or tB3 and tB4, are, apart from dead times, located symmetrically.

[0088] During each of the 510 periods, the SB sequence is in one of the N or P states. The N and P states alternate in successive 510 periods. The 510 periods are separated by 520 periods during which the sequence is in the O state.

[0089] Because the duty cycles of SB1 and SB2 are equal, the periods 520 have identical durations. Furthermore, because the duty cycles of SB1 and SB2 are equal to 0.5, the SB sequence is such that the periods 510 have identical durations for both states N and P. It follows that the states N, O, and P of the SB sequence are arranged symmetrically with respect to time tBS.

[0090] The fact of predicting that the SA sequences ( figure 3 ) and SB ( figure 4 ) have their states, respectively N and P, and N, O and P, arranged symmetrically with respect to the respective instants tAS and tBS allows, in operation, to avoid the presence of a direct current component (I135, figure 1 ) in transformer 130 ( figure 1 ).

[0091] In alternative embodiments, any other values ​​for the duty cycles and / or phase shifts of cycles SA1, SA2 and SB1, SB2 of the respective sequences SA and SB can be provided to guarantee the absence of such a DC component. However, predicting SA and SB sequences that are symmetrical with respect to times tAS and tBS makes it simpler to avoid the DC component. Furthermore, this advantageously results in the transformer current exhibiting symmetrical evolutions in both directions of flow, which simplifies obtaining a modeled value of the transformer current as a function of time and a modeled power P such as that defined in relation to the figure 2 .

[0092] Although the O state of the SB sequence results, in the example of the SB1 and SB2 cycles described above, from the phase shift of the SB1 and SB2 cycles relative to each other, alternatively, the O state of the SB sequence can be obtained in a different way than described above, for example: by replacing, at time tB1, the switching of cycle SB2 from high to low by switching cycle SB1 from low to high and, at time tB2, switching cycle SB1 from low to high by switching cycle SB2 from high to low (dotted line 522); and / or by replacing, at time tB3, the switching of cycle SB2 from low to high by switching cycle SB1 from high to low and, at time tB2, switching cycle SB1 from high to low by switching cycle SB2 from low to low (dotted line 524).

[0093] The SB switching sequence, described above in its application to the H 120 bridge, can be applied to the H 110 bridge ( figure 1 ) similarly, by replacing, with respect to the application to bridge 120 described above, branches 121 and 122 respectively with branches 111 and 112 ( figure 1 ). More specifically, for this, compared to the application to bridge 120 described above, we replace the switches T21H, T21L, T22H and T22L respectively with the switches T11H, T11L, T12H and T12L.

[0094] THE figures 5 And 6 represent, schematically and in the form of timing diagrams, examples of implementation methods for steps in a control process for a converter of the type of converter 100 of the figure 1 For example, these steps are implemented in step 220 ( figure 2 ) calculation of the switching times ti of the bridges and generation of the control signals to be applied to the bridges. These examples do not take into account the switching dead times.

[0095] Each of the figures 5 And 6 represents patterns as a function of time t normalized by the cutting period Tc: of an S110 switching sequence of the H 110 bridge; of an S120 switching sequence of the H 120 bridge; of the V135 voltage ( figure 1 ) across the terminals of the leakage inductance 135 of the transformer 130; and a current I135 flowing in the leakage inductance (that is, in the example shown in figure 1 , in winding 131 of the transformer).

[0096] At each stage of the figures 5 And 6 , the SA and SB sequences of figures 3 et 4 are used, and the switching times tA1, tA2, tB1, tB2, tB3 and tB4 of the SA and SB sequences are calculated. These times correspond to the times ti described in relation to the figure 2 .

[0097] In the examples of figures 5 And 6 , the voltage V1 is positive, and applied in a given direction (between nodes 141 and 142, figure 1 ) to a state P of the S110 sequence, and in the opposite direction (between nodes 142 and 141) to a state N of the S110 sequence. In other words, the N and P states of the S110 sequence correspond to the respective directions of application of the voltage V1 to the transformer by the H-bridge 110. A zero voltage is applied between nodes 141 and 142 (in other words, these nodes are short-circuited) to a state O of the S110 sequence.

[0098] Similarly, the voltage V2 is positive and applied in a given direction (between nodes 151 and 152, figure 1 ) to a state P of the S120 sequence, and in the opposite direction (between nodes 142 and 141) to a state N of the S120 sequence. In other words, the N and P states of the S120 sequence correspond to the respective directions of application of the voltage V2 to the transformer by the H-bridge 120. A zero voltage is applied between nodes 151 and 152 to a state O of the S120 sequence.

[0099] There figure 5 illustrates a first mode of operation or control of the converter. More specifically, the figure 5 is a reproduction of figure referenced 6A from the aforementioned patent and patent applications.

[0100] In figure 5 For example, it is assumed that the voltage V2 across the H120 bridge, multiplied by the turns ratio n (i.e., a voltage value of n*V2), is greater than the peak value of the voltage V1 across the H110 bridge. In other words, the voltage value n*V2 is greater than that of the voltage V1. Put another way, the ratio V1 / V2 between the voltages across the H110 bridge and the H120 bridge, respectively, is less than the turns ratio n between winding 132 on the H120 bridge side and winding 131 on the H110 bridge side. Furthermore, it is assumed that energy flows from the H110 bridge to the H120 bridge.

[0101] Put another way, in the example of the figure 5 , the 100 converter ( figure 1 ) is considered to be operating in boost mode and the power transferred from bridge 110 to bridge 120 is considered positive.

[0102] In the first mode of operation, the SA sequence described above in relation to the figure 3 is applied to the bridge at H110, repeatedly at the switching frequency. Thus, the S110 sequence corresponds to the SA sequence. In other words, the P and N states of the S110 sequence correspond to the respective P and N states of the SA sequence, and the S110 sequence does not take the O state. Furthermore, the SB sequence described above in relation to the figure 4 is applied to the bridge in H120, repeatedly at the switching frequency. Thus, the S120 sequence corresponds to the SB sequence. In other words, the P, O, and N states of the S110 sequence correspond to the respective P, O, and N states of the SB sequence.

[0103] In the first mode of operation, the input and output switching of a given state of sequence SB between states N and P occurs at different states of sequence SA. In the present step, the input switching tB2 and output switching tB3 of state P of sequence SB occur at states P and N of sequence SA, respectively. The input switching tB4 and output switching tB1 of state N of sequence SB occur at states N and P of sequence SA, respectively. In other words, a switching (here at time tA2) of sequence SA occurs between the input and output switching (times tB2, tB3) of state P of sequence SB. A switching (here at time tA1) of sequence SA occurs between the input and output switching (times tB4, tB1) of state N of sequence SB.

[0104] The switching instants tB1, tB2, tB3, and tB4 of the SB sequence are defined relative to the instants tA1 and tA2 of the SA sequence by two parameters x and y. The parameters x and y range from 0 to 0.5 and each corresponds to a fraction of the switching cycle time Tc (the inverse of the switching frequency). The duration y*Tc (represented by the letter y) separates each instant tB2 from the following instant tA2, and the duration x*Tc (represented by the letter x) separates each instant tA2 from the following instant tB3. In other words, the parameters x and y represent the differences (or durations) between the switching instants, respectively tA2 and tB3, and tB2 and tA2. The parameters x and y are calculated so that the instants ti are those for which the modeled value allows the transfer of a power equal to the setpoint P*, for example so that the modeled value of transmitted power is equal to the setpoint value.

[0105] The SA and SB sequences are then generated, or applied to bridges 110 and 120, based on the calculated x and y parameters. The SA and SB sequences have a phase shift dϕ between time tAS and time tBS. The sequences are then applied to bridges H 110 and 120.

[0106] Preferably, to generate the SA and SB sequences, a reference time for the sequences is defined. This time is generated, for example, by a clock signal at the switching frequency. As an example, the reference time is, in this case, the time tA1 when the S110 sequence applied to the H110 bridge transitions to state P. The other switching times for the SA and SB sequences are defined by: a phase shift φ 1 of the S110 sequence between the reference instant tA1 of the sequences and the instant tAS of symmetry of the SA sequence applied to the bridge at H 110; a phase shift φ 2 of the S120 sequence between the reference instant tA1 of the sequences and the instant tBS of symmetry of the SB sequence applied to the bridge in H 120; a duty cycle D1 of the S110 sequence, defined by the ratio between the duration of a state P of the S110 sequence, and the cycle time Tc; and a duty cycle D2 of the S120 sequence, defined by the ratio between the duration of a state P of the S120 sequence, and the cycle time Tc.

[0107] Phase shifts φ 1 and φ 2, and the cyclic ratios D1 and D2, are given by the following equalities: D 1 = 0 , 5 φ 1 = π 2 D 2 = x + y φ 2 = 2 . π . 1 2 + x − y 2

[0108] The voltage V135 takes the values ​​-V1-n*V2, -V1, V1-n*V2, n*V2-V1, V1, and V1+n*V2 when the respective states of the sequences S110 and S120 are, respectively, N and P, N and O, P and P, N and N, P and O, and P and N, namely, respectively, between the instants tA2 and tB3, tB3 and tB4, tB2 and tA2, tB4 and tA1, tB1 and tB2, and tA1 and tB1.

[0109] Preferably, to calculate the parameters x and y, we seek the values ​​of these parameters which allow us to obtain an equality between the i0 values ​​of the current I135 in the transformer at times tB1 and tA2. In other words, the parameters x and y, and therefore, from these parameters, the times tA1, tA2, tB1, tB2, tB3 and tB4, result from calculations based on a desired, or targeted, equality between the i0 values ​​of the current I135 in the transformer at times tB1 of the sequence SB and tA2 of the sequence SA, the times tB1 and tA2 being separated by the time tB2 of the sequence SB.

[0110] For this, we use a model of the converter, in the manner described in relation to the figure 2 The model provides modeled values ​​of the current I135 at times tB1 and tA2 as a function of the parameters x and y, taking into account values ​​such as those of the voltages V1 and V2 across bridges 110 and 120. The converter model can be any standard model of a converter comprising two H-bridges connected by a transformer. An example of a converter model is described in relation to the figure 8 of the aforementioned requests, and corresponds to one or more algebraic expressions providing the modeled value of the current I135 as a function of time t and parameters x and y.

[0111] Preferably, the calculation of parameters x and y consists of choosing, from all possible values ​​of parameters x and y, those for which the current I135 has the same modeled values ​​at times tB1 and tA2. In other words, the calculated values ​​of parameters x and y are those for which a relation of equality between the modeled values ​​as a function of parameters x and y is satisfied or verified. This can be obtained by any standard method for finding parameters for which a relation between values ​​as a function of these parameters is satisfied.

[0112] As indicated and described in the aforementioned patent and patent applications, the relationship between function values ​​of parameters x and y can be algebraic, and the choice of values ​​of parameters x and y that satisfy this relationship can be made by choosing only one of the two values ​​(for example that of parameter x) and calculating the other of the two values ​​from this algebraic relationship, or can be calculated numerically, the procedure for finding parameters x and y then typically being a numerical search by successive iterations.

[0113] Because the current I135 exhibits symmetrical evolutions in its two directions of current flow, the desired equality between values ​​i0 of the current I135 at times tB1 and tA1 also corresponds to a desired equality between values ​​-i0 of the current I135 at times tA1 and tB3 separated by the switching time tB2, as well as to desired equalities, in absolute value, of the current I135 at consecutive times (i.e. not separated by a switching) tA1 and tB1, and / or at consecutive times tA2 and tB3.

[0114] In alternative configurations, the calculation of parameters x and y can be performed based on any desired relationship defined as a function of the current in the transformer at the switching instants of the two sequences. An example of such a desired relationship is a desired equality between a ratio of current values ​​at times tB1 and tA2, and a predefined value that may be different from 1.

[0115] From the converter model, in particular from the modeled values ​​I135 and V135, we can calculate a modeled value P of the power transferred as a function of time by the converter from the H110 bridge to the H120 bridge, on average at each repetition of the switching sequences.

[0116] Preferably, the calculation of the parameters x and y then consists of choosing, among all possible values ​​of the parameters x and y, those for which the power setpoint P* described previously is equal to the modeled value P. In other words, the calculation of the parameters x and y is based on an equality between the power setpoint P* and the power calculated from the converter model and the values ​​of the voltages V1 and V2.

[0117] In examples of calculating the parameters x and y, the model corresponds to an algebraic expression P(x, y) providing the value P as a function of the parameters x and y. The calculation can then include any usual method, for example numerical, of solving the equation P* = P(x, y) to obtain a set of values ​​to choose for the parameters x and y.

[0118] In other examples, the converter model is digital, and the parameters x and y are calculated numerically by any usual method of solving the equation, such as an iterative method.

[0119] More preferably, the calculation of the parameters x and y, and therefore of the switching times of the SA and SB sequences, is based both on the desired equality between currents at times tB1 and tA2, and on the desired equality between the power setpoint P* and the power transferred by the converter between the bridges.

[0120] In some variations, the power setpoint can be replaced by any representative value of a setpoint supplied to the converter, such as a setpoint for a current to be drawn from one of the bridges and / or supplied by the other bridge. The desired equality between the setpoint P* and the modeled power is then replaced by an equality between this current setpoint and an average value of this current, modeled according to the converter model, over each repetition of the SA and SB sequences.

[0121] There figure 5 has been described for an example where the voltages V1 and V2 are positive. However, the voltages V1 and V2 can be algebraic quantities. When the voltages V1 and / or V2 are negative, the description given above of the figure 5 applies by replacing the values ​​of voltages V1 and V2 with their absolute values, and, for example, by swapping the N and P states of the respective sequences S110 and / or S120.

[0122] There figure 6 illustrates a second mode of operation or control of the converter. More specifically, the figure 6 corresponds to figure referenced 7A of the aforementioned patent and patent applications, in which the references x' and y' have been replaced by the respective references x and y.

[0123] Similarly, as with the figure 5 , there figure 6 illustrates an example where the voltages V1 and V2 have positive values, however, in a manner similar to what was indicated in relation to the figure 5 , the description given below of the figure 6 applies to negative values ​​of voltages V1 and / or V2.

[0124] In figure 6 , as in figure 5 The voltage value n*V2 is greater than that of the voltage V1, and it is further predicted that energy will flow from the bridge at H110 to the bridge at H120. In other words, in the example of the figure 6 , the 100 converter ( figure 1 ) is considered to be operating in boost mode and the power transferred from bridge 110 to bridge 120 is considered positive.

[0125] As in the stage of the figure 5 , the SA and SB sequences described above in relation to the figures 3 et 4 are applied to the H 110 and 120 bridges, repeatedly at the switching frequency.

[0126] However, in the second operating mode, the input and output switching of a given state of sequence SB between states N and P occurs at the same state of sequence SA. In the present step, the input switching tB2 and output switching tB3 of state P of sequence SB occur at state N of sequence SA. The input switching tB4 and output switching tB1 of state N of sequence SB occur at state P of sequence SA.

[0127] The switching times tB1, tB2, tB3, and tB4 of sequence SB are determined relative to the times tA1 and tA2 of sequence SA by two parameters x and y. Parameters x and y range from 0 to 0.5 and each corresponds to a fraction of the switching cycle time Tc. The duration y*Tc (represented by y) separates each time tB2 from the following time tB1, and the duration x*Tc (represented by x) separates each time tB3 from the following time tA2.

[0128] The parameters x and y are, for example, calculated in a similar way to that described in relation to the figure 5 , namely, more preferably: based on an equality between modeled values, calculated from the converter model and from values ​​of the voltages across the bridges, of a current in the transformer at switching instants of the SA and SB sequences (for example the instants tA1 and tB3); and / or based on an equality between the power represented by the setpoint P* and the corresponding modeled power value P, calculated from the converter model and from values ​​of the voltages V1 and V2 across the bridges.

[0129] The SA and SB sequences are then generated from the x and y parameters obtained and applied to the bridges at H 110 and 120. The SA and SB sequences have the phase shift dϕ, described above, between the time tAS and the time tBS.

[0130] Preferably, the switching times of the SA and SB sequences are defined relative to the reference time tA1 by the phase shifts φ 1 and φ 2 and the cyclic ratios D1 and D2 defined in relation to the figure 5 .

[0131] Phase shifts φ 1 and φ 2, and the duty cycles D1 and D2, are given from the parameters x and y by the following equalities: D 1 = 0 , 5 φ 1 = π 2 D 2 = y φ 2 = 2 . π . 1 2 − x − y 2

[0132] The voltage V135 takes the values ​​-V1, V1-n*V2, n*V2-V1, and V1 when the respective states of sequences S110 and S120 are, respectively, N and O, P and P, N and N, P and O, namely, respectively, between times tB3 and tB4, tB2 and tA2, tB4 and tA1, tB1 and tB2. The respective states of sequences S110 and S120 are also, respectively, N and O, P and O between times, respectively, tB1 and tA1, tB3 and tA2.

[0133] Thus, in this example of the second operating mode, compared to the example of the first operating mode of the figure 5 , we replaced the respective states N and P, P and N of the sequences S110 and S120 between the times tA1 and tB1 ( figure 5 ) and tA2 and tB3 ( figure 5 ), by the states N and O, P and O. This comes from the fact that the state P of the sequence SB starts after time tA1 and ends before time tA2, and that the state N of the sequence SB starts after time tA2 and ends before time tA1.

[0134] In this example, it follows that, between the times tA1 and tB1 of the figure 6 , the current I135 varies in the opposite direction to the direction of variation of the current I135 between the instants tB1 and tA1 of the figure 5 Similarly, between times tB3 and tA2 of the figure 6 , the current I135 varies in the opposite direction to the direction of variation of the current I135 between times tA2 and tB3 of the step of the figure 5 .

[0135] As a result, the power transferred by the converter is relatively low in the second operating mode and relatively high in the first operating mode. Therefore, the power setpoint P* determines whether the converter is controlled according to the first or second operating mode.

[0136] As an example, for a given converter model, the transferable power P in the first operating mode, in the case described in relation to the figure 5 , is given by the following equation: P = V 1 2 4 . m . fsw . L − 4 − 4 . m 2 . x 2 + 4 . m 2 − 2 . m + 2 . x + m − m 2 with fsw the repetition or switching frequency, L a value of the 135 leakage inductance of the converter and m a ratio between a ratio of the voltages across the two bridges and a transformation ratio of the transformer, in this example equal to V1 / (V2.n).

[0137] Again, as an example, for the same given converter model, the transferable power P in the second operating mode, in the case described in relation to the figure 6 , is given by the following equation: P = V 1 2 4 . fsw . L 1 − 2 . x 1 − m . 1 − 2 . x − 4 . x

[0138] For the two equations [Math 3] and [Math 4] above, the equalities of current values ​​at two instants tB1 and tA2 as described in relation to the figures 5 And 6 are respected.

[0139] The two equations [Math 3] and [Math 4] above do not take into account possible switching dead times, for example determined so as to ensure zero voltage switching (ZVS) of the converter switches.

[0140] As an example, consider an operation where the switching frequency fsw is fixed prior to the calculation of the parameters x and y described previously for each of the first and second operating modes. The fixed value of the frequency fsw is, for example, called the nominal value of the switching frequency (or the repetition frequency of the sequences described previously).

[0141] As an example, the nominal value of the switching frequency can be determined in the manner indicated in the aforementioned patent and patent applications, for example be constant or be calculated from the values ​​of the voltages V1 and V2 and the power setpoint P*.

[0142] As described in relation to the figure 8 of the aforementioned patent and patent applications, for this nominal switching frequency value, the modeled value according to a converter model of the power transferred P by the converter from the H 110 bridge to the H 120 bridge can then be determined as a function of the parameter x for each of the two operating modes, for example from equations [Math 3] and [Math 4].

[0143] Depending on a power setpoint P*, it is then determined which of the first and second operating modes is used and the value of the parameter x for this operating mode, based on the equality between the setpoint P* and the modeled value of transferable power P.

[0144] In such an example, when the converter is operating in the second mode and the power setpoint P* increases beyond a maximum power that the second operating mode can transfer, the converter switches from the second mode to the first operating mode. Similarly, when the converter is operating in the first mode and the power setpoint P* decreases below a minimum power that the first operating mode can transfer, the converter switches from the first mode to the second operating mode. Switching between the first and second operating modes is described, for example, in the aforementioned patent and patent applications, in connection with their figures 8 , 9 and 10.

[0145] In an ideal case without dead time, at the nominal value of the switching frequency, the switching from the first to the second operating mode, or from the second to the first operating mode, takes place in zero current switching (ZCS), when the parameter x of equations [Math 3] and [Math 4] is equal to 0.

[0146] However, in practice, switching downtime is expected. Determining the value or duration of each switching downtime is within the capabilities of a person skilled in the art, for example, based on the description of figure 9 of the aforementioned patent and patent applications.

[0147] However, in the converter model determined above by equations [Math 3] and [Math 4], these switching dead times are not taken into account and result in the zero voltage switching (ZVS) being lost during a switch between the first and second operating modes implemented for x equal to 0, at the nominal switching frequency value.

[0148] It is thus proposed, in a switching converter of the type of that of the figure 1 that is to say comprising two H-bridges connected by a transformer, and controlled according to the first and second modes described above, to switch between the first operating mode at the nominal value of the switching frequency and the second operating mode at the nominal value of the switching frequency taking into account dead times, for example so as to maintain ZVS operation when switching from one to the other of these two operating modes, for example by ensuring a ZVS margin.

[0149] To achieve this, the switching frequency value is adjusted during the switchover between these two operating modes, based on a duration D equal to the duration of the switching dead times or equal to the product of the dead time duration and a zero-voltage switching margin, or ZVS margin. In other words, the converter is controlled according to one of the two operating modes and at the nominal switching frequency value before switching between the two operating modes; the converter is controlled according to the other of the two operating modes and at the nominal switching frequency value after switching between the two operating modes; and the switching frequency value is adjusted during the switchover between the two operating modes based on the duration of the dead times or based on the duration of the dead times and a ZVS margin.

[0150] There figure 7 represents, schematically and in block form, a method of adapting the value of the switching frequency fsw during a transition between the two operating modes described previously.

[0151] At step 700 (block "CALCULATE Plim1 FOR MODE 2 AT fsw0 WITH D AND Plim2 FOR MODE 1 AT fsw0 WITH D"), the limit value Plim1 of the modeled power P at which zero-voltage switching is lost at the nominal value fsw0 of the switching frequency fsw is calculated for the second operating mode. Furthermore, the limit value Plim2 of the modeled power P at which zero-voltage switching is lost at the nominal value fsw0 of the switching frequency fsw is calculated for the first operating mode. At this step, a duration D, for example, equal to the duration of the dead time, is taken into account. In other words, at this step, the switching dead time is considered.

[0152] For a given value of the switching frequency fsw and taking into account the duration D of the dead times, the limiting value xlim of the parameter x for which the voltage switching to zero is lost in each of the two operating modes is equal to the product of the duration D and the switching frequency fsw. In other words: xlim = D . fsw

[0153] Put another way, this means that the duration xlim / fsw which the parameter x represents is then equal to the duration D.

[0154] Thus, in each of the first and second modes of operation, at the value fsw0 of the frequency fsw, the zero-voltage switching is lost if x is less than xlim = D.fsw0.

[0155] For example, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the power Plim1 in the second operating mode is obtained by setting the parameter x to the value xlim1 = D*fsw0 in the model of the converter's second operating mode, that is, here in equation [Math 4]. In other words: Plim 1 = V 1 2 4 . fsw 0 . L 1 − 2 . xlim 1 1 − m 1 − 2 . xlim 1 − 4 . xlim 1

[0156] For example, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the power Plim2 in the first operating mode is obtained by setting the parameter x to the value xlim2 = D*fsw0 in the model of the converter's first operating mode, that is, here in equation [Math 3]. In other words: Plim 2 = V 1 2 4 . m . fsw 0 . L − 4 − 4 . m 2 . xlim 2 2 + 4 . m 2 − 2 . m + 2 . xlim 2 + m − m 2

[0157] Given that the value of the parameter m, the nominal value fsw0 of the switching frequency fsw, the duration D, the value V1 and the value L of the leakage inductance are known, it is therefore possible to calculate the limit power values ​​Plim1 and Plim2.

[0158] In step 702 (block "CALCULATE Plim FOR MODE 1 and MODE 2 WITH D=0"), the limiting value Plim of the modeled power P that the converter is capable of supplying for each of the first and second modes at the nominal switching frequency fsw0 and for a zero value of the parameter x is calculated. In other words, the Plim power is a limit power transferable at the frequency fsw0 in an ideal case without dead times. As an example, step 702 is implemented after step 700, although it can also be implemented before or simultaneously with step 700.

[0159] For example, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the power Plim is obtained by setting the parameter x to a value of zero in the model of the first or second operating mode of the converter, that is, here in equation [Math 3] or [Math 4]. In other words: Plim = V 1 2 4 . m . fsw 0 . L m − m 2 = V 1 2 4 . fsw 0 . L 1 − m

[0160] At step 704 (block "CALCULATE fsw1 for MODE 2 AT Plim WITH D AND fsw2 FOR MODE 1 AT Plim WITH D"), a value fsw1 for the switching frequency fsw is calculated. The value fsw1 is the frequency fsw value that, in the second operating mode and when the parameter x is equal to xlim, transfers the power Plim calculated in step 702. Similarly, at this step 704, a value fsw2 for the switching frequency fsw is calculated. The value fsw2 is the frequency fsw value that, in the first operating mode and when the parameter x is equal to xlim, transfers the power Plim calculated in step 702.

[0161] In an approximate solution, it is assumed that at step 704 the value xlim is constant and equal to D times fsw0. The solution is called an "approximate solution" because, in practice, for a given frequency fsw, the limiting value xlim for which zero-voltage switching is lost is equal to D.fsw. However, during the transition between the two operating modes, the value of the frequency fsw varies and does not remain equal to the nominal value fsw0, which leads to a variation in the value xlim. In the approximate solution, it is nevertheless assumed that, during the frequency adaptation during the transition between the two operating modes, the value xlim is constant and equal to D times fsw0.

[0162] As an example, for the approximate solution, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the frequency value fsw1 for the second operating mode is obtained by setting the parameter x to the value xlim1 = D.fsw0 and the power P to the value Plim in the model of the converter's second operating mode, i.e., here in equation [Math 4]. In other words: fsw 1 = V 1 2 4 . Plim . L 1 − 2 . xlim 1 1 − m 1 − 2 . xlim 1 − 4 . xlim 1

[0163] As an example, for the approximate solution, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the frequency value fsw2 for the first operating mode is obtained by setting the parameter x to the value xlim2 = D.fsw0 and the power P to the value Plim in the model of the converter's first operating mode, i.e., here in equation [Math 3]. In other words: fsw 2 = V 1 2 4 . m . Plim . L − 4 − 4 . m 2 . xlim 2 2 + 4 . m 2 − 2 . m + 2 . xlim 2 + m − m 2

[0164] In an exact solution, between the powers Plim1 and Plim2, that is to say during the transition between the two modes of operation during which the frequency is adapted, we consider that the value xlim is not constant and equal to D times fsw0, but on the contrary that it varies with the frequency according to the relation xlim = D.fsw.

[0165] Thus, in the exact solution, calculating the frequency fsw1 for which the second operating mode allows transferring Plim with x equal to xlim1, that is to say D times fsw1, amounts to calculating, with the model of transferable power P of the second operating mode, what is the value xlim1 of the parameter x allowing to transfer the power Plim calculated in step 702, then to deduce the value fsw1 of frequency fsw on the basis of the relation fsw1 = xlim1 / D. Similarly, in the exact solution, calculating the frequency fsw2 for which the first operating mode allows Plim to be transferred with x equal to xlim2, i.e. D times fsw2, amounts to calculating, with the transferable power model P of the first operating mode, what is the value xlim2 of the parameter x allowing the transfer of the power Plim calculated in step 702, and then deducing the value fsw2 of the switching frequency fsw on the basis of the relation fsw2 = xlim2 / D.

[0166] As an example, for the exact solution, when the converter operates as a boost converter and the power transferred from bridge 110 to bridge 120 is positive, the value xlim1 that allows transferring the power Plim in the second mode without losing the zero-voltage switching is obtained by solving a quadratic equation where xlim1 = fsw1.D is the unknown. This quadratic equation is obtained from the converter model for the second operating mode, i.e., equation [Math 4], by fixing the power P in this model to the value Plim and replacing fsw with fsw1, i.e., with xlim1 / D. This amounts to solving the following quadratic equation: 0 = − 4 . m + 8 . xlim 1 2 + 4 . m − 6 − 4 . Plim . L D . V 1 2 . xlim 1 + 1 − m

[0167] By asking: alim 1 = − 4 . m + 8 , blim 1 = 4 . m − 6 − 4 . Plim . L / D . V 1 2 , And clim 1 = 1 − m , we obtain: xlim 1 = − blim 1 − 4 . alim 1 . clim 1 2 . alim 1

[0168] We then obtain fsw1 = xlim1 / D.

[0169] As another example, for the exact solution, when the converter operates as a boost converter and the power transferred from bridge 110 to bridge 120 is positive, the value xlim2 required to transfer the power Plim in the first operating mode without losing the zero-voltage switching is obtained by solving a quadratic equation where xlim2 is the unknown. This quadratic equation is derived from the converter model for the first operating mode, i.e., equation [Math 3], by setting the power P in this model to the value Plim and replacing fsw with fsw2, i.e., with xlim2 / D. Thus, this amounts to solving the following quadratic equation: 0 = − 4 . m 2 − 4 . xlim 2 2 + 4 . m 2 − 2 . m − 4 . m . Plim . L D . V 1 2 . xlim 2 + m − m 2

[0170] By asking: alim 2 = − 4 . m 2 − 4 , blim 2 = 4 . m 2 − 2 . m − 4 . m . Plim . L / D . V 1 2 , And clim 2 = m − m 2 , we obtain: xlim 2 = − blim 2 − 4 . alim 2 . clim 2 2 . alim 2

[0171] We then obtain fsw2 = xlim2 / D.

[0172] As an example, steps 700, 702 and 704 are implemented in step 220 of the figure 2 .

[0173] As an example, steps 700, 702 and 704 are implemented by a converter processing circuit, for example by processing circuit 180 ( figure 1 ) of the converter 100, for example by the processing unit of this circuit 180, for example adapting the program stored in the memory of circuit 180.

[0174] There figure 8 represents, schematically and by means of curves, a method of implementing yet another control step of the converter of the figure 1 More specifically, the figure 8 illustrates the evolution of the fsw frequency as a function of the transferable power P during a transition between the first operating mode (MODE 1 in figure 8 ) to the nominal value fsw0 of the fsw frequency and the second operating mode (MODE 2 in figure 8 ) to the nominal value fsw0 of the frequency fsw. In other words, the figure 8 illustrates a converter control step according to an embodiment where the fsw frequency is adapted during a transition between the two operating modes, based on the duration D.

[0175] As can be seen on the figure 8 , at the nominal value fsw0 of the frequency fsw, when the transferable power P is less than the power Plim1 calculated in step 700 ( figure 7 ), the converter is controlled according to the second operating mode, MODE 2. In this first region, the value of the parameter x and the corresponding times ti are calculated, for example, in step 220 ( figure 2 ), then the corresponding control signals are applied, for example, to the bridge switches in step 230 ( figure 2 ).

[0176] In this first operating region, the parameter x is calculated in the manner indicated previously in relation to step 220 ( figure 2 ), based on an equality between the power setpoint P* and the modeled transferable power P, the latter being determined for example by equation [Math 4] in a given model of the converter.

[0177] For example, at each variation of the power setpoint P*, as long as the setpoint P* remains less than the value Plim1, the parameter x is recalculated to respect the equality P* = P, and the switching instants ti are deduced or calculated from the recalculated value of the parameter x.

[0178] When, in the second mode MODE 2, the frequency fsw is at its nominal value and the setpoint P*, therefore the power P, reaches the limit value Plim1, the parameter x is then equal to xlim1 = D.fsw0. From the power Plim1 onwards, the value of the frequency fsw is adjusted and is no longer maintained constant and equal to fsw0.

[0179] More specifically, in MODE 2, in a second operating region where the power P is between Plim1 and Plim, the value of the frequency fsw is between fsw0 and fsw1. For example, in this second region, the value of the frequency fsw decreases as the power P increases.

[0180] In this second region, the value of the parameter x, the value of the frequency fsw and the corresponding instants ti are calculated, for example, in step 220 ( figure 2 ), then the corresponding control signals are applied, for example, to the bridge switches in step 230 ( figure 2 ) at the frequency fsw. In particular, in this second region, the value of the parameter x is equal to D.fsw0 (approximate solution) or to D.fsw (exact solution).

[0181] For example, in MODE 2, for a setpoint P* (therefore a modeled power P) equal to a value P1 between Plim1 and Plim, the corresponding value f1 of the frequency fsw is calculated using the model of the second mode, and by setting the parameter x to the value x1 equal to fsw0.D (approximate solution) or to f1.D (exact solution).

[0182] As an example, for the approximate solution, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the frequency fsw value f1 for the second operating mode is obtained by setting the parameter x to the value x1 = D.fsw0 = xlim1 and the power P to the value P1 in the model of the converter's second operating mode, i.e., here in equation [Math 4]. In other words, this amounts to solving the following equation: f 1 = V 1 2 4 . P 1 . L 1 − 2 . x 1 1 − m 1 − 2 . x 1 − 4 . x 1

[0183] As an example, for the exact solution, when the converter operates as a boost converter and the power transferred from bridge 110 to bridge 120 is positive, the value x1 required to transfer power P1 in the second mode without losing the zero-voltage switching is obtained by solving a quadratic equation where x1 is the unknown. This quadratic equation is derived from the converter model for the second operating mode, i.e., equation [Math 4], by setting the power P in this model to the value P1 and replacing fsw with f1, i.e., with x1 / D. This amounts to solving the following quadratic equation: 0 = − 4 . m + 8 . x 1 2 + 4 . m − 6 − 4 . Plim . L D . V 1 2 . x 1 + 1 − m

[0184] By asking: a 1 = − 4 . m + 8 , b 1 = 4 . m − 6 − 4 . P 1 . L / D . V 1 2 , And c 1 = 1 − m , we obtain: x 1 = − b 1 − 4 . a 1 . c 1 2 . a 1

[0185] We then obtain f1 = x1 / D.

[0186] When, in the second mode MODE 2, the setpoint P*, therefore the power P, reaches the limit value Plim, the parameter x is then equal to D.fsw0 (approximate solution) or to D*fsw1 (exact solution), and the frequency fsw is at the value fsw1.

[0187] At this point, the converter's control or operating mode switches from the second mode, MODE 2, to the first mode, MODE 1. For this to happen, the frequency value fsw switches from fsw1 to fsw2. In other words, switching from MODE 2 to MODE 1 involves a jump in the frequency fsw from its value fsw1 to its value fsw2. In practice, the value fsw2 is higher than the value fsw1.

[0188] In MODE 1, in a third operating region where the power P is between Plim and Plim2, the value of the frequency fsw is between fsw2 and fsw0. For example, in this third region, the value of the frequency fsw decreases as the power P increases.

[0189] In this third region, the value of the parameter x, the value of the frequency fsw and the corresponding instants ti are calculated, for example, in step 220 ( figure 2 ), then the corresponding control signals are applied, for example, to the bridge switches in step 230 ( figure 2 ) at the frequency fsw. In particular, in this third region, the value of the parameter x is equal to D.fsw0 (approximate solution) or to D.fsw (exact solution).

[0190] For example, in MODE 1, for a setpoint P* (therefore a modeled power P) equal to a value P2 between Plim and Plim2, the corresponding value f2 of the frequency fsw is calculated using the model of the first mode, and by setting the parameter x to the value x2 equal to fsw0.D (approximate solution) or to f2.D (exact solution).

[0191] As an example, for the approximate solution, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the frequency fsw value f2 for the first operating mode is obtained by setting the parameter x to the value x2 = D.fsw0 = xlim2 and the power P to the value P2 in the model of the converter's first operating mode, i.e., here in equation [Math 3]. In other words, this amounts to solving the following equation: f 2 = V 1 2 4 . m . P 2 . L − 4 − 4 . m 2 . x 2 2 + 4 . m 2 − 2 . m + 2 . x 2 + m − m 2

[0192] As an example, for the exact solution, when the converter operates as a boost converter and the power transmitted from bridge 110 to bridge 120 is positive, the value x2 required to transfer power P2 in the first mode without losing the zero-voltage switching is obtained by solving a quadratic equation where x2 is the unknown. This quadratic equation is derived from the converter model for the second operating mode, i.e., equation [Math 4], by setting the power P in this model to the value P2 and replacing fsw with f2, i.e., with x2 / D. This amounts to solving the following quadratic equation: 0 = − 4 . m 2 − 4 . x 2 2 + 4 . m 2 − 2 . m + 2 − 4 . m . P 2 . L D . V 1 2 . x 2 + m − m 2

[0193] By asking: a 2 = − 4 . m 2 − 4 , b 2 = 4 . m 2 − 2 . m + 2 − 4 . m . P 2 . L / D . V 1 2 , And c 2 = m − m 2 , we obtain: x 2 = − b 2 − 4 . a 2 . c 2 2 . a 2

[0194] We then obtain f2 = x2 / D.

[0195] When, in MODE 1, the setpoint P*, and therefore the power P, reaches the limit value Plim2, the parameter x is then equal to xlim2 = D.fsw0, and the frequency fsw is at its value fsw0. In a fourth operating region where the power P is greater than Plim2, the value of the frequency fsw is no longer suitable and is maintained constant and equal to fsw0.

[0196] A transition from MODE 2 to MODE 1 was described above in relation to the figure 8 In particular, during this transition, the switching from MODE 2 to MODE 1 when the power P equals Plim involves a jump in the frequency fsw from the value fsw1 to the value fsw2. A person skilled in the art can deduce from this description, by symmetry of operation, the description of a transition from MODE 1 to MODE 2, for example, following a decrease in the power setpoint P*. Specifically, during a transition from MODE 1 to MODE 2, the switching from MODE 1 to MODE 2 when the power P equals Plim involves a jump in the frequency fsw from the value fsw2 to the value fsw1.

[0197] There figure 9 represents, schematically and in block form, one implementation method for the ordering step of the figure 8 .

[0198] At a step 900 (block "P ≤ Plim1"), it is checked if the power setpoint P* (i.e. the power P) is less than the value Plim1.

[0199] If this is the case (output Y of block 900), the converter is in the first operating region and is controlled according to MODE 2. Furthermore, the process continues to step 902 (block "fsw = fsw0") where the frequency fsw is set to its nominal value fsw0, and then the process loops back to step 900. In the first region, the parameter x and the times ti are, for example, calculated as indicated in relation to the figure 8 .

[0200] If this is not the case (output N of block 900), the process continues to a next step 904 (block "Plim1 < P < Plim").

[0201] At step 904, it is checked whether the power setpoint P* (i.e. the power P) is between Plim1 and Plim.

[0202] If this is the case (output Y of block 904), it means that the converter is in the second operating region described previously and is therefore controlled according to MODE 2. The process then continues to step 906 (block "fsw ≠ fsw0") where the frequency fsw is no longer kept constant and equal to its nominal value fsw0, but is adapted based on the duration D, that is, taking into account the duration of the dead times. In this second region, the parameter x and the times ti are, for example, calculated in the manner indicated in relation to the figure 8 Step 906 is followed by step 904.

[0203] If this is not the case (output N of block 904), the process continues at a step 908 (block "P ≥ Plim").

[0204] At step 908, it is checked whether the power setpoint P* (i.e., the power P) is greater than or equal to Plim.

[0205] If this is not the case (output N of block 908), because step 908 follows step 904, it means that the power P is neither between Plim1 and Plim, nor greater than Plim, and therefore the power P is less than Plim1 and the converter is in the first operating region. The process then continues to step 900 as illustrated in figure 9 , or, alternatively, at step 902.

[0206] If this is the case (output Y of block 908), the process continues to step 910 (block "MODE = MODE 1") during which the converter switches from operating mode MODE 2 to operating mode MODE 1 as illustrated by dashed lines in figure 9 Step 910 is followed by step 912 (block "Plim ≤ P < Plim2") as shown in figure 9 or, alternatively, a step 918 (block "P ≥ Plim2).

[0207] At step 912, it is checked whether the power setpoint P* (i.e., the power P) is between Plim and Plim2.

[0208] If this is the case (output Y of block 912), it means that the converter is in the third operating region described previously and is therefore controlled according to MODE 1. The process then continues to step 9914 (block "fsw ≠ fsw0") where the frequency fsw is no longer kept constant and equal to its nominal value fsw0, but is adapted based on the duration D, that is, taking into account the duration of the dead times. In this third region, the parameter x and the times ti are, for example, calculated in the manner described in relation to the figure 8 Step 914 is followed by step 912.

[0209] If this is not the case (output N of block 912), the process continues at a step 916 (block "P < Plim").

[0210] At step 916, it is checked if the power setpoint P* (i.e., the power P) is less than Plim.

[0211] If this is not the case (output N of block 916), because step 916 follows step 912, it means that the power P is neither between Plim and Plim2, nor less than Plim, and therefore that the power P is greater than Plim2 and the converter is in the fourth operating region. The process then continues to step 918 as illustrated in figure 9 , or, alternatively, at a step 920 (block "fsw = fsw0").

[0212] If this is the case (output Y of block 916), the process continues to step 922 (block "MODE = MODE 2") during which the converter switches from operating mode MODE 1 to operating mode MODE 2 as illustrated by the dashed lines in figure 9 Step 922 is followed by step 904, as shown in figure 9 , or, alternatively, from step 900.

[0213] At step 918, it is checked whether the power setpoint P* (i.e., the power P) is greater than or equal to the value Plim2.

[0214] If this is the case (output Y of block 918), the converter is in the fourth operating region and is controlled according to MODE 1. Furthermore, the process continues to step 920 where the frequency fsw is set to its nominal value fsw0, and then the process loops back to step 920. In the fourth region, the parameter x and the times ti are, for example, calculated as indicated in relation to the figure 8 .

[0215] If this is not the case (output N of block 918), the process continues at step 912.

[0216] In the example above, it was assumed that when the power P is equal to Plim, the converter is controlled according to MODE 1. A person skilled in the art will be able to adapt the above procedure to examples where, when the power P is equal to Plim, the converter is controlled according to MODE 2. More generally, a person skilled in the art will be able to modify the above example by replacing strict inequalities with simple inequalities (covering the equal case) and vice versa.

[0217] We described previously, in relation to the figures 7 à 9 An approximate solution is used to calculate Plim, the frequency fsw when the power P is between Plim1 and Plim, and the frequency fsw when the power P is between Plim and Plim2. The parameter x is considered constant and equal to D.fsw0. When the duration D is equal to the duration of the dead time and the power P is between Plim1 and Plim, this amounts to overestimating the value xlim = D.fsw because the frequency fsw is then between fsw0 and fsw1, and is therefore less than fsw0. Conversely, when the duration D is equal to the duration of the dead time and the power P is between Plim and Plim2, this amounts to underestimating the value xlim = D.fsw because the frequency fsw is then between fsw2 and fsw0, and is therefore greater than fsw0.

[0218] Thus, at least with the approximate solution, instead of using a duration D equal to the duration of the dead times, it is preferable to use a duration D equal to the product of the duration of the dead times by a margin ZVS, denoted A and strictly greater than 1, for example between 1.1 and 1.5, for example equal to 1.1. This is particularly true for the calculation of the power Plim2, the frequency fsw2, and for the calculation of the value of the frequency fsw when the power P is between Plim and Plim2.

[0219] In the description given above in relation to the figures 5 à 9 The converter operates in boost mode, meaning that the voltage V1 is less than a voltage n.V2 (m<1), and the power transferred from bridge 110 to bridge 120 is positive. A person skilled in the art can adapt the description given in relation to the figures 5 à 9to cases where the converter operates in buck mode, meaning that the voltage V1 is greater than the voltage n.V2 (m>1), and where the power transferred from bridge 110 to bridge 120 is positive. For this, a person skilled in the art can, for example, refer to the French patent and the EP and US patent applications mentioned previously or, as an alternative example, make the following changes to the variables in the calculations described above: The voltage V1 described in relation to a boost operation is replaced by a voltage n.V2 in a buck operation, with V2 the voltage across the bridge 120 in buck mode, and the voltage V2 described in relation to a boost operation is replaced by a voltage V1 / n in a buck operation, with V1 the voltage across the bridge 110 in buck mode.

[0220] Furthermore, whether the converter operates in step-down or step-up mode, it has been assumed until now that the power transferred from bridge 110 to bridge 120 is positive. However, a person skilled in the art can adapt the above description to the case where the power transferred from bridge 110 to bridge 120 is negative, either because the calculations with negative power are similar, or by considering that the transferred power is always positive until the calculation of the phase shift, or phase difference, dϕ between sequences SA and SB, and by choosing the sign of the transferred power at the time of calculating this phase shift.

[0221] A particular example of a converter model has been described previously, this model being determined by equations [Math 3] for MODE 1 and [Math 4] for MODE 2. A person skilled in the art is able, from this example, to adapt the calculation steps described above to other converter models, for example taking into account resistances which would lead to a model determined by equations different from equations [Math 3] and [Math 4].

[0222] Various embodiments and variations have been described. A person skilled in the art will understand that some features of these various embodiments and variations could be combined, and other variations will become apparent to a person skilled in the art.

[0223] Finally, the practical implementation of the described methods and variants is within the reach of the person in the trade, based on the functional indications given above.

Claims

1. Method of controlling a converter (100) comprising two H bridges (110, 120) coupled by a transformer (130), each of the two bridges comprising two branches in parallel between two terminals of the bridge, each branch comprising two switches in series, wherein: - repetitions, at a repetition frequency (fsw), of two sequences (SA, SB) of switching operations between a plurality of states (P, O, N) are respectively applied to the two bridges; - a first (SB) of the two sequences is applied to a first of the two bridges, and said plurality of states of the first sequence comprises a first state (N) in which two switches of the first bridge disposed on one side of a first of the two terminals of the first bridge are respectively ON and OFF, a second state (P) in which said two switches of the first bridge are respectively OFF and ON, and a third state (O) in which two switches of the first bridge disposed on one side of a same terminal among the two terminal of the first bridge are OFF and the other two switches of the first bridge are ON; - a second sequence (SA) of the two sequences is applied to a second of the two bridges, and said plurality of states of the second sequence comprises a first state (N) in which two switches of the second bridge disposed on one side of a first of the two terminals of the second bridge are respectively ON and OFF, and a second state (P) in which said two switches of the second bridge are respectively OFF and ON; - a voltage is applied to the transformer by the first bridge in a given direction at the first state (N) of the first sequence (SB), and in an opposite direction to the given direction at the second state (P) of the first sequence (SB); - in a first control mode (MODE 1), switching operations to enter and leave a given state of the first sequence (SB) among the first and second states of the first sequence (SB) occur in different states of the second sequence (SA); - in a second control mode (MODE 2), switching operations to enter and leave a given state of the first sequence (SB) among the first and second states of the first sequence (SB) occur in a same state of the second sequence (SA); and the method being characterized in that: - each transition between the first mode (MODE 1) at a nominal value (fsw0) of the repetition frequency (fsw) and the second mode (MODE 2) at the nominal value (fsw0) of the repetition frequency (fsw) comprises an adaptation of the repetition frequency (fsw) determined by a first duration (D) equal to a switching dead time duration or to the product of the switching dead time duration and of a zero-voltage switching margin.

2. Method according to claim 1, wherein each transition comprises, based on a model of the converter (100): - the calculation (700) of a first power (Plim2) transferable between the two bridges (110, 120) with the first mode (MODE 1) and the nominal value (fsw0) of the repetition frequency (fsw), the first power (Plim2) being determined by the first duration (D); - the calculation (700) of a second power (Plim1) transferable between the two bridges (110, 120) with the second mode (MODE 2) and the nominal value of the repetition frequency (fsw), the second power (Plim1) being determined by the first duration (D); - the calculation (702) of a third power (Plim) transferable between the two bridges (110, 120) in the first and second modes (MODE 1, MODE 2) at the nominal value (fsw0) of the repetition frequency (fsw) and without taking into account the first duration (D); - the calculation (704) of a first value (fsw2) of the repetition frequency (fsw) in the first mode (MODE 1), the first value (fsw2) of the repetition frequency being determined by the third transferable power (Plim) and the first duration (D); and - the calculation (704) of a second value (fsw1) of the repetition frequency (fsw) in the second mode (MODE 2), the second value (fsw1) of the repetition frequency being determined by the third transferable power (Plim) and the first duration.

3. Method according to claim 2, wherein: each transition from the first mode (MODE 1) to the second mode (MODE 2) comprises a jump from the first value (fsw2) to the second value (fsw1) of the repetition frequency (fsw); and each transition from the second mode (MODE 2) to the first mode (MODE 1) comprises a jump from the second value (fsw1) to the first value (fsw2) of the repetition frequency (fsw).

4. Method according to claim 2 or 3, wherein: each transition from the first mode (MODE 1) with the nominal value (fsw0) of the repetition frequency (fsw) to the second mode (MODE 2) with the nominal value (fsw0) of the repetition frequency (fsw) successively comprises: - in the first mode (MODE 1), transiting from the first power (Plim2) to the third power (Plim) by transiting from the nominal value (fsw0) of the repetition frequency (fsw) to the first value (fsw2) of the repetition frequency (fsw), - at the third power (Plim), switching from the first mode (MODE 1) to the second mode (MODE 2) by switching from the first value (fsw2) of the repetition frequency (fsw) to the second value (fsw1) of the repetition frequency, and - in the second mode (MODE 2), transiting from the third power (Plim) to the second power (Plim1) by transiting from the second value (fsw1) of the repetition frequency (fsw) to the nominal value (fsw0) of the repetition frequency (fsw); and each transition from the second mode (MODE 2) with the nominal value (fsw0) of the repetition frequency (fsw) to the first mode (MODE 1) with the nominal value (fsw0) of the repetition frequency (fsw) successively comprises: - in the second mode (MODE 2), transiting from the second power (Plim1) to the third power (Plim) by transiting from the nominal value (fsw0) of the repetition frequency to the second value (fsw1) of the repetition frequency, - at the third power (Plim), switching from the second mode (MODE 2) to the first mode (MODE 1) by switching from the second value (fsw1) of the repetition frequency (fsw) to the first value (fsw2) of the repetition frequency (fsw), and - in the first mode (MODE 1), transiting from the third power (Plim) to the first power (Plim2) by transiting from the first value (fsw2) of the repetition frequency to the nominal value (fsw0) of the repetition frequency (fsw).

5. Method according to any of claims 2 to 4, wherein: the switching operations between the states of each of the first and second sequences occur at times (tA1, tA2, tB1, tB2, tB3, tB4) resulting from calculations based on an equality between: - a set point power to be transferred between the bridges (110, 120) by the converter (100); and - a transferable power (P) calculated based on the converter model and based on values of the voltages (V1, V2) across the bridges (110, 120).

6. Method according to claim 5, wherein the set point is calculated as a function of a voltage value received by one of the bridges (110) and / or of a voltage value to be delivered by the other of the bridges (120).

7. Method according to claim 6, wherein the set point is calculated so that the converter has a PFC-type operation.

8. Method according to any of claims 5 to 7, wherein said calculations are further based on a desired equality between values (i0) of a current (I135) in the transformer (130) at one of the switching times of one of the two sequences and at one of the switching times of the other one of the two sequences.

9. Method according to any of claims 5 to 8, wherein: the model comprises a first parameter (x) representative of a second duration between switching times of the two sequences (SA, SB) in the first mode (MODE 1) and of a third duration between switching times of the two sequences (SA, SB) in the second mode (MODE 2); the first transferable power (Plim2) in the first mode (MODE 1) is calculated based on the model by using a first value of the first parameter (x) corresponding to an equality of the first and second durations; the third transferable power (Plim) in the first and second modes (MODE 1, MODE 2) is calculated based on the model by using a zero value of the first parameter; and the second transferable power (Plim1) in the second mode (MODE 2) is calculated based on the model by using the first value of the first parameter (x) corresponding to an equality of the first and third durations.

10. Method according to claim 9, wherein: in the first mode (MODE 1), the model is determined by the following equation: P = V 1 2 4 . m . fsw . L − 4 − 4 . m 2 . x 2 + 4 . m 2 − 2 . m + 2 . x + m − m 2 where P is the transferable power, V1 is a voltage across one of the two bridges, x is the first parameter, fsw is the repetition frequency, L is a value of a leakage inductance of the converter, and m is a ratio of a ratio of the voltages across the two bridges to a transformation ratio of the transformer; and in the second mode (MODE 2), the model is determined by the following equation: P = V 1 2 4 . fsw . L 1 − 2 . x 1 − m . 1 − 2 . x − 4 . x .

11. Device (180) configured to implement a method according to any of claims 1 to 10.

12. Converter (100) comprising a device according to claim 11.