Pulse generator and circuit for controlling a power electronics component

Abstract

One aspect of the invention relates to a pulse generator (4) for generating a current pulse on an output branch (46) upon receipt of a trigger command, comprising: • a first half-bridge (431, 432) and a second half-bridge (433, 434), the latter comprising bidirectional switches; • an inductor (45) connected between the two half-bridges; and • a controller configured, according to different configurations of the half-bridges, to charge the inductor (45) and discharge it into the output branch (46).

Description

DESCRIPTION TITLE: PULSE GENERATOR AND CONTROL CIRCUIT OF A POWER ELECTRONIC COMPONENT TECHNICAL FIELD OF THE INVENTION

[0001] The technical field of the invention is that of pulse generation, for example to control a power electronic component. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0002] Power electronics components are electronic components configured to control the flow of a high-intensity electric current. Examples include metal-oxide-semiconductor field-effect transistors (MOSFETs) and high-electron-mobility transistors (HEMTs). The current to be controlled flows between the source and drain of the transistor. Current control is achieved by applying a current or voltage to the transistor's gate.

[0003] With the development of power components, there is a need to be able to control these components via a transmission line. For example, a power component might be placed on a characterization bench located away from the control circuit. The transmission line connecting the control circuit to the component can be several tens of centimeters, or even a few meters, long. These electrical characteristics will therefore influence the control signal generated by the source.

[0004] Therefore, there is a need to provide a means that can generate an arbitrary pulse towards a load (such as a power component) at high frequency and through a transmission line, regardless of the length of the latter (even zero length).

[0005] The document ["Multilevel Transmission Line Pulse (MTLP) tester" T. Daenen et al., 2004 Electrical Overstress / Electrostatic Discharge Symposium, Grapevine, Texas, USA, 2004, pp. 1-6] describes a device for generating an arbitrary pulse across a transmission line. This device generates two pulses of different amplitudes and durations which, when superimposed within the transmission line, can enable the switching of a power component.

[0006] The document [“Snapback Device Studies Using Multilevel TLP and Multiimpedance TLP Testers,” E. Grund, 2005 Electrical Overstress / Electrostatic Discharge Symposium, Anaheim, CA, USA, 2005, pp. 1–9] describes two devices for generating an arbitrary pulse without requiring data processing. In the first example, the transmission line between the voltage source and the component is modified to incorporate an impedance that reflects part of the incident wave. The superposition of the incident and reflected waves creates an overshoot at the component, i.e., a pulse followed by a plateau. In the second example, a high-impedance resistor is placed on the transmission line, after the electronic component. This impedance performs the same function, namely, using the superposition of the incident and reflected waves to generate a pulse and the plateau.

[0007] However, these devices require systematic modification of the transmission line and a prior study of the position and impedance value before being used to control the reflection of incident and reflected waves. Furthermore, modifying the shape of the control signal (amplitude and / or duration) requires modifying the transmission line.

[0008] The paper [“Gate Lifetime of P-Gate GaN HEMT in Inductive Power Switching,” B. Wang et al., IEEE 2023 35th International Symposium on Power Semiconductor Devices and ICs (ISPSD), June 2023] describes a device for generating a pulse on the gate of a power electronic component, specifically a HEMT. The device forms a gate loop with the HEMT. The gate loop includes an inductor connected to a voltage source and a fast transistor short-circuiting the HEMT gate. The transistor, when conducting, charges the inductor; and, when blocked, discharges the inductor onto the HEMT gate.

[0009] The document [“Dynamic Gate Breakdown of p-Gate GaN HEMTs in Inductive Power Switching”, B. Wang & al., IEEE Electron Device Letters, Vol. 44, No. 2, February 2023] describes a device similar to the aforementioned device.

[0010] The document [“A Current Source Gate Drive Achieving Switching Loss Savings and Gate Energy Recovery at 1 -MHz”, W. Eberle et al., IEEE Trans, on Power Electronic, Vol. 23, No 2, March 2008] discloses a current source making it possible to This system generates an arbitrary pulse to a power component while controlling the maximum amplitude of the current applied to the component's gate. The current source comprises four unidirectional switches and an inductor, all five elements connected in an H configuration. Controlling the switches allows the inductor to be charged and the stored energy delivered to the component's capacitive load as a high, constant current. SUMMARY OF THE INVENTION

[0011] The invention solves at least partially the problems mentioned above, by allowing the generation of an arbitrary current pulse, which can be used to control a power component through a transmission line that can be up to several meters long.

[0012] A transmission line is defined as a set of two conductors (for example, the core and sheath of a coaxial cable or two conductors twisted together) carrying an electrical signal from a source (or transmitter or generator) to a load (or receiver). The usual criterion that a transmission line has a length of the same order of magnitude as, or greater than, the wavelength of the highest frequency of the signal to be transmitted is not a determining criterion in the case of this invention. The transmission line can therefore have an arbitrary length, from a few centimeters to several meters, or even be zero in length.

[0013] The invention relates to an arbitrary current pulse generator on an output branch upon receipt of a trigger command, comprising: first, second, third and fourth terminals configured to be connected to, respectively, first, second, third and fourth electrical potentials V1, V2, V3, V4; first, second, third and fourth switches; an inductor; the output branch; and a controller.

[0014] The first switch is connected between the first terminal and a first node; the second switch is connected between the second terminal and the first node; the third switch is connected between the third terminal and a second node; the fourth switch is connected between the fourth terminal and the second node; and the inductor is connected between the first node and the second node; and the output branch is connected to the second node.

[0015] The pulse generator is notable in that the third and fourth switches are bidirectional switches and in that the controller is also configured to: charge the inductance by closing the first and fourth switches and opening the second and third switches; and upon receiving the trigger command, discharge the inductance into the output branch by opening the fourth switch.

[0016] A unidirectional switch is understood to be a switch that systematically passes through a current with a determined direction of flow.

[0017] A bidirectional switch is a switch that can block the flow of current, regardless of its direction of flow.

[0018] Charging the inductor means storing a quantity of energy in the inductor.

[0019] Closing the first and fourth switches charges the inductor, allowing it to store a certain amount of energy. The inductor is thus charged between potentials V1 and V4. The charge is considered positive when V1 > V4.

[0020] Opening the fourth switch connects the charged inductor between the first terminal and the output branch. The inductor then injects at least some of its stored energy into the output branch as a current. Discharging the inductor provides a brief, high-amplitude current pulse. When the output branch is connected to a transmission line (for example, via a control circuit), the generator allows current to be injected into the transmission line, thus controlling a component. of the power connected to it. This injection allows, for example, the switching of the component.

[0021] The switches connected to the second node are bidirectional. They therefore block any current leakage from the second node except through the output branch. Thus, a substantial portion, or even all, of the current produced by the inductor is injected into the output branch, regardless of the electrical characteristics of the transmission line to which it may be connected. Indeed, the switches may exhibit small leakage currents, which we will not consider here.

[0022] Contrary to the invention, the generator described in the document "A Current Source Gate Drive Achieving Switching Loss Savings and Gate Energy Recovery at 1 MHz" [W. Eberle et al., IEEE Trans. on Power Electronic, Vol. 23, No. 2, March 2008] includes unidirectional switches connected to the second node. These switches can cause a current flow far exceeding the leakage current and reduce the efficiency of the command sent to the power component. Furthermore, the current magnitude depends on the load connected to the generator and / or the electrical properties of the transmission line.

[0023] Advantageously, the closing of the first and fourth switches is carried out when the controller receives a load command.

[0024] Advantageously, the controller is also configured to, after the fourth switch is opened, close the third switch and open the first switch.

[0025] Thus, after current is injected, the third terminal is connected to the output branch, which is held at the third potential, V3. It is advisable that this third potential, V3, be equal to the holding potential of the component being controlled. Holding potential, or holding voltage, refers to a potential difference that, when applied to the component's gate, keeps it in a specific state, for example, conducting or blocking.

[0026] Advantageously, the third switch is closed upon receipt of an end command. By end command, we mean a command indicating the cessation of current pulse generation.

[0027] According to one development, the second switch comprises a bidirectional switch and a diode, with the bidirectional switch and the diode connected in parallel between the second terminal and the first node. The diode conducts, for example, a current flowing from the second terminal to the first node.

[0028] Thus, when the third switch is closed, the inductor continues to discharge due to the current flowing through the diode of the second switch. When the voltage across the inductor is zero, the diode of the second switch prevents further (positive) charging of the inductor.

[0029] Advantageously, the controller is also configured to: negatively charge the inductance by closing the second and third switches and opening the first and fourth switches; and discharge the inductance again into the output branch by opening the third switch, preferably upon re-reception of the trigger signal.

[0030] Negative charging means charging the inductor with a current direction opposite to the direction of the current when the first and fourth switches are open. Unlike the previous steps, the current flows from the third terminal to the second terminal. When the third switch opens, the negatively charged inductor is connected between the second terminal and the output branch. The current injected into the output branch is negative compared to the previously injected current. Discharging the inductor provides a brief, negative current pulse. When the output branch is connected to a transmission line, the generator allows current to be injected into the transmission line, thus controlling a connected power component. This injection allows, for example, switching the component from the conducting state to the blocking state (whereas the previous switching allowed switching from blocking to conducting).

[0031] According to one development, the controller is also configured to, after the third switch is opened, open the second switch and close the fourth switch.

[0032] In this way, the fourth terminal is connected to the output branch, and the latter is maintained at the fourth potential, V4. This potential is, for example, equal to 0 V, i.e., connected to ground. The component can thus be stored in a blocked state (if the third potential V3 allows a conducting state to be maintained). When the electrical potentials satisfy the following inequality: V3 - V2 > |V1 - V4|, they allow the inductor to charge negatively faster than it charges positively. Since V4 is limited by its holding function, V2 can be chosen as large as necessary.

[0033] Advantageously, the first switch also includes a bidirectional switch and a diode, the bidirectional switch and the diode being connected in parallel between the first node and the first terminal, the diode of the first switch and the diode of the second switch being conducting to a current flowing in the same direction between the first terminal and the second terminal.

[0034] Thus, when the fourth switch is closed, the inductance automatically discharges into the diode of the first switch and remains discharged afterwards.

[0035] Advantageously, the third and fourth switches each comprise two transistors, for example metal / oxide / semiconductor (or MOS) or high electron mobility (or HEMT), mounted in common sources.

[0036] Advantageously, the pulse generator includes an impedance to reduce the amplitude of a bounce from the output branch.

[0037] When the output branch is connected to a transmission line, variations in impedance or the component's input impedance can reflect some of the injected current, known as "bounce." Impedance reduction helps protect the generator from bounce, for example, from a transmission line. It also allows the generator's output to be matched so that the bounce is dampened and returned to the transmission line with a reduced or even zero amplitude. This results in a reduction or elimination of multiple reflections in the transmission line. Thus, the signal received by the component at the end of the line is clearly defined. The risk of unintentional overvoltage at the component's input is avoided.

[0038] Indeed, the reduction is all the more effective when the impedances are inserted close to a fixed potential (for example, a ground). Small parasitic bounces can be observed when the impedances they traverse are separated by a fixed potential distance less than the wavelength characteristic of these small bounces. In practice, a distance of a few A few centimeters or less is sufficient. Although reflections do not affect the operation of devices, they introduce noise into the signals and prevent the accurate reading of certain characterization information, such as rise time. Therefore, minimizing the distance between an impedance and a fixed potential improves reflection reduction.

[0039] Consider an example where the generator is connected to a controlled impedance transmission line, which is itself connected to a load. Only the first switch is closed, allowing the inductor to discharge into the output branch. We can define a branch as having, connected in series, a first ground point, a power supply connected between ground and the first terminal, the first switch, the inductor, the output branch, the transmission line, the grid of the power component, and a second ground point.

[0040] Inserting an impedance into the circuit at the end of the controlled impedance line can therefore, to effectively reduce the amplitude of a bounce off that branch, be done at a distance from a ground point that is less than the wavelength of the current that can travel along that branch (on the order of a few centimeters, or even less than a centimeter, to a few meters). In practice, inserting this impedance between the second node and the output branch is sufficient because when the bounce returns to the second node, one of the third or fourth switches is closed and connects the second node to a fixed potential. The distance between the impedance and a fixed potential (for example, ground) is then very short, allowing the bounce to be absorbed effectively.

[0041] Advantageously, the pulse generator includes a first current limiter connected in series with the inductor when the first, second, third and fourth switches are in a configuration allowing the inductor to be loaded.

[0042] Switch configuration refers to the configuration of each switch with respect to its open or closed state.

[0043] The current limiter allows the maximum current flowing through the inductor to be limited when the inductor is loaded.

[0044] For example, the first limiter is connected in series with the fourth switch between the second node and the fourth terminal.

[0045] Advantageously, the pulse generator includes a second current limiter connected in series with the third switch between the second node and the third terminal.

[0046] Advantageously, the pulse generator includes a power supply configured to apply the first, second, third and fourth electrical potentials V1, V2, V3, V4 on, respectively, the first, second, third and fourth terminals, with V1 > V4 and V3 > V2.

[0047] Advantageously, the power supply is configured to apply electrical potentials such that V1 - V4 > |V2 - V3|. This allows for rapid positive charging of the inductor to generate a positive current on the output branch (i.e., flowing out of the output branch). Since V3 is limited by its holding function, V1 can be chosen as large as required.

[0048] The power supply can be configured to apply electrical potentials such that V3 - V2 > |V1 - V4|. This allows for a rapid negative charge of the inductor to generate a negative current on the output branch (i.e., entering the output branch). Since V4 is also limited by its holding function, V2 can be chosen as large as required.

[0049] Alternatively, the power supply is configured to apply electrical potentials such that V1 - V4 = |V2 - V3|. This allows the inductor to charge positively as quickly as it charges negatively.

[0050] Advantageously, the third potential V3 is approximately equal to the high holding potential of the component being controlled. By approximately equal, we mean equal to within 20%, or even 10%. Similarly, the fourth potential V4 can be approximately equal to the low holding potential of the component being controlled.

[0051] The invention also relates to a control circuit for a power electronic component through a transmission line, the control circuit being notable in that it comprises: a pulse generator according to the invention; a switch; and a controller.

[0052] The switch includes: an output terminal configured to be connected to the transmission line; a first input terminal and a second input terminal, the first input terminal of the switch being configured to be connected to the output branch of the pulse generator, the second input terminal of the switch being configured to be connected to a first voltage source; and a first bidirectional switch and a second bidirectional switch, the first switch being connected between the first input terminal and the output terminal, the second switch being connected between the second input terminal and the output terminal.

[0053] The controller is configured to: send a trigger command to the pulse generator; upon sending the trigger command to the pulse generator, close the first switch and open the second switch; and an instant after sending the trigger command to the pulse generator or after closing the first switch, open the first switch and close the second switch.

[0054] The control circuit allows the pulse generator to be controlled to prepare a current pulse and send it to a transmission line (and to a component that can be connected to it). The circuit is unique in that the current pulse is sent in near-superposition with the voltage supplied by the first voltage source. Near-superposition means that the current pulse and the voltage supplied by the source do not coincide directly but alternate continuously.

[0055] The switch's bidirectional switches allow for an exclusive connection to be established between the pulse generator and the output terminal, and between the first voltage source and the output terminal. Thus, there is never a connection between the pulse generator and the first voltage source. Therefore, there is no current leakage from the pulse generator to the voltage source (even if the Bidirectional switches can have parasitic parallel capacitances that allow a small, high-frequency current to pass through. In other words, the entire current generated by the pulse generator is injected towards the output terminal and thus into a transmission line that can be connected to it. The switching of the component is therefore performed efficiently.

[0056] Advantageously, the controller is also configured to send a charge command to the pulse generator before sending the trigger command. This charge command enables the generator's inductance to be charged (positively or negatively). The inductance's charge level between charging and the triggering of the pulse generation depends on the time elapsed between sending the two commands and also on the potential difference applied to the inductance (i.e., V1 - V4 or V3 - V2). For a given potential difference, the longer the time between sending the two commands, the greater the amount of energy stored in the inductance. The resulting current pulse will therefore have a higher amplitude. Conversely, the shorter the time between sending the two commands, the less energy is stored in the inductance.The current pulse will therefore have a moderate amplitude. It is thus possible, by adjusting the time between sending the two commands, to control the amplitude of the generated current pulse.

[0057] Advantageously, the controller is further configured to send a termination command to the pulse generator no later than when the first switch is opened. This sets the generator's output branch to a constant potential (in this case, V3). Furthermore, pulse generation is interrupted. The pulse duration can therefore be controlled independently of the switch closures. Moreover, this prevents disconnecting the load (transmission line and component) from the output branch, which could damage the inductor and / or a component connected to the transmission line.

[0058] This also allows the transmission line potential to be maintained at the third potential V3, which, when non-zero, enables a continuous and differentiable transition when the first voltage source takes over. This reduces parasitic potential variations and the risk of inducing unwanted switching of the component.

[0059] Advantageously, the control circuit includes a first impedance to reduce the amplitude of a bounce from the transmission line at the first and second input terminals. The first impedance is designed to match the line and the control circuit so that the bounce is damped. In this way, the current sent to the components through the transmission line, which is reflected back to the control circuit, is not sent back to the electronic component. The same control signal is therefore not received multiple times by the component, preventing spurious switching. Furthermore, the first impedance protects the control circuit, including the switch, as well as the components at the switch's input, such as the first voltage source or the pulse generator.

[0060] Let's define a first branch comprising, connected in series when the first switch is closed, a first ground point, the pulse generator, the switch, the transmission line, the power component grid, and a second ground point. Let's also define a second branch comprising, connected in series when the second switch is closed, a third ground point, the first voltage source, the switch, the transmission line, the power component grid, and the second ground point. The transmission line, the component grid, and the second ground point are part of the two aforementioned branches. Inserting an impedance in each of the branches or an impedance in the common portion of the branches can help reduce the reflections propagating within these branches.

[0061] Advantageously, the first impedance is connected between the first and second switches and the output terminal of the switch. In this way, the first impedance is inserted into the common portion of the circuit where a wave propagation is possible. Furthermore, this first impedance is positioned at a distance of less than a few centimeters from the first and third ground points. This single impedance is sufficient to protect both the switch switches and the components connected to the switch's input.

[0062] Advantageously, the control circuit also includes a second impedance. The first impedance is connected between the output terminal and the first input terminal, preferably between the first input terminal and the first switch, and the second impedance is connected between the output terminal and the second input terminal, preferably between the second input terminal and the second switch. In this way, each impedance is part of a separate branch and allows for the protection of an input element of the switch, whether it be the first voltage source or the pulse generator. Furthermore, this method allows for different impedance values, enabling optimal matching of each branch. This improves the amplitude reduction of a bounce.

[0063] Advantageously, the first impedance is connected between the output terminal and the first switch. This also protects the first switch. As a reminder, the switches can be semiconductor devices such as MOSFETs or HEMTs. Following the same principle, the second impedance is preferentially connected between the output terminal and the second switch.

[0064] Advantageously, the first impedance is connected between the first switch and the first input terminal. This reduces the distance from the first impedance to ground, improving rebound damping. This, in turn, better protects the pulse generator. Following the same principle, the second impedance is preferably connected between the second switch and the second input terminal.

[0065] It is advantageous to combine the addition of impedances at the different positions mentioned above because this allows the voltage to be distributed across each impedance.

[0066] In one implementation, the controller is configured to vary the amplitude of the voltage delivered by the first voltage source over time, between a first amplitude and a second amplitude different from the first amplitude. The controller is further configured so that the variation between the first and second amplitudes occurs when the second switch is open. Thus, the amplitude change takes place when the first The voltage source is disconnected from the output terminal. Therefore, this voltage variation does not add to the current pulse injected at the output terminal. Consequently, there is no modulation of the current pulse resulting from the modulation of the current delivered by the first source. Furthermore, the first voltage source allows the application of different voltages that correspond respectively to the holding voltages of a component in its various states, for example, blocked or conducting.

[0067] According to one development of this embodiment, the controller is configured to modulate the first and / or second amplitude delivered by the first voltage source as a function of time. In this way, the control circuit generates a signal with an arbitrary shape that may include first and second modulations separated by a current pulse.

[0068] According to another development of this embodiment, the controller is also configured to close the second switch before the current pulse is generated.

[0069] Thus the first amplitude of the voltage delivered by the first voltage source can be applied to the output terminal, before sending the current pulse, and the second amplitude can be applied after sending the current pulse.

[0070] Alternatively, the switch further includes: a third input terminal configured to be connected to a second voltage source; and a third bidirectional switch, connected between the third input terminal and the output terminal.

[0071] The controller is also configured to: before sending the trigger command to the pulse generator, close the second switch and open the first and third switches; as soon as the trigger command is sent to the pulse generator, open the second switch and close the first switch; and One moment after sending the trigger command to the pulse generator, close the third switch and open the first and second switches.

[0072] This embodiment does not require a first voltage source with a modulating output voltage. It therefore simplifies circuit fabrication. Furthermore, the first and second sources can be fixed and constant voltage sources. In this way, the voltage sources allow the holding voltages to be applied to the component. The transition from the first holding voltage to the second holding voltage occurs when the third switch closes and the second switch opens, both switches being open for a period of time sufficient to allow the current pulse from the pulse generator to pass through without being reduced or modulated.

[0073] At least one of the voltage sources can be modulated.

[0074] It is also advantageous for the switch to include additional input terminals and additional switches, preferably bidirectional. This allows for the connection of additional voltage sources or generators.

[0075] In this development, the control circuit may further include a pulse generator according to the additional invention, and the switch includes: a fourth input terminal, configured to be connected to the output branch of the additional pulse generator; and a fourth bidirectional switch connected between the fourth input terminal and the output terminal, the controller being configured to: send a trigger command to the additional pulse generator; upon sending the trigger command to the additional pulse generator, close the fourth switch and open the second switch and, if applicable, the third switch; and one instant after sending the trigger command to the additional pulse generator or after closing the fourth switch, open the fourth switch and close the second switch or, if necessary, the third switch.

[0076] The invention and its various applications will be better understood by reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES

[0077] The figures are shown for illustrative purposes only and are not intended to limit the invention. Unless otherwise specified, the same element appearing in different figures has a unique reference numeral.

[0078] Figure 1 schematically presents a first embodiment of a control circuit according to the invention, as well as a transmission line and a capacitive load.

[0079] Figure 2 schematically presents a second embodiment of the control circuit according to the invention.

[0080] Figure 3 shows a timing diagram that can be implemented by the control circuit of Figure 2.

[0081] Figure 4 schematically presents a third embodiment of the control circuit according to the invention.

[0082] Figure 5 schematically presents an embodiment of a pulse generator according to the invention.

[0083] Figure 6 schematically presents the pulse generator of Figure 5 in several operating modes. DETAILED DESCRIPTION

[0084] Figure 1 presents a first embodiment of a control circuit 1 according to the invention. The control circuit 1 is designed to control (also called "drive") a power electronic component (represented by dashed lines), and more particularly its gate. The gate of the power component is represented in Figure 1 by a capacitive load 2. This is a common example of gate modeling. The equivalent capacitance can typically range from 0.5 nF to several hundred nanofarads.

[0085] The capacitive load 2 is controlled via a transmission line 3. This embodiment is applicable, for example, to the characterization of power electronic components. The components can be placed on a characterization bench (also called a "tester") located away from the control circuit 1 used to operate the components. The control circuit 1 is connected to the components to be characterized by means of the transmission line 3. This step can be performed automatically, using a robotic arm that connects the transmission line 3 sequentially to each component. The transmission line may therefore have bends or be subject to tensile forces that can vary between the different components.

[0086] The control circuit 1 according to the invention allows each component to be controlled independently of the length of the control line 3 or its geometric or electrical characteristics, provided that the equivalent impedance of this line 3 is known. The transmission line 3 can be a coaxial cable or a pair of conductors twisted together. It can measure from a few centimeters to several meters. By extrapolating the physical principles involved, the control circuit 1 could offer the same technical effects and advantages even if the transmission line 3 were of zero length.

[0087] The impedance of transmission line 3 can typically be between 10 Q and 100 Q.

[0088] In the embodiment of Fig. 1, the control circuit 1 comprises a switch 6, a pulse generator 4, and a controller (not shown). The control circuit 1 is connected to a first voltage source 51, in this case a modulating one, and to the transmission line 3. In the development of the control circuit 1, the first voltage source 51 is part of the circuit 1.

[0089] The pulse generator 4 is responsible for generating current pulses for the component, in order to control its switching. The first voltage source 51 can be used to deliver a modulating voltage V51, alternating between a first voltage, called the "low holding voltage," and a second voltage, called the "high holding voltage," before and after switching, respectively. These low and high holding voltages allow, for example, to maintain the component in a given state (e.g., low or high, blocked or conducting) by applying holding voltages. The controller is responsible for modulating the holding voltage delivered by the first voltage source 51. The switch 6 and the controller are responsible for connecting the pulse generator 4 or the first voltage source 51 to the transmission line 3 in order to switch the component and maintain it in a given state.

[0090] The switch 6 includes a first input terminal 611, a second input terminal 612 and an output terminal 62. The output terminal 62 of the switch 6 is connected to the transmission line 3. The first input terminal 611 of the switch 6 is connected to an output 46 of the pulse generator 4. The second input terminal 612 of the switch 6 is connected to the first voltage source 51.

[0091] The controller is, for example, an electronic circuit, a programmable calculator, or a computer. It could be, for example, a Texas Instruments™ LaunchXL-F28379D™ board or a programmable electronics device (called an "FPGA" for "Field-Programmable Gate Array") instrumented on a Red Pitaya™ board.

[0092] It includes means of communication and / or control for communicating at least with the switch 6 and the pulse generator 4. For example, it is configured to send a trigger command to generate a current pulse to the pulse generator 4. It can also send, prior to the trigger command, a load command to the pulse generator 4, so that the latter prepares to send a current pulse to the switch 6. It can also control the switch 6 so that one or both, or neither, of its input terminals 611, 612 is connected to the output terminal 62 (and thus to the transmission line 3).

[0093] In an alternative implementation (not shown), generator 4 can be connected to switch 6 by means of an additional transmission line. This allows generator 4 to be located further away from switch 6 and facilitates the integration of control circuit 1. Furthermore, several generators 4 may be available, so using a transmission line between generators 4 and switch 6 allows the generator 4 connected to switch 6 to be changed.

[0094] Figure 2 schematically presents a second embodiment of a control circuit 1 according to the invention, including in particular a variant implementation of the switch 6.

[0095] In this embodiment, the switch 6 comprises a first switch 631 and a second switch 632. The first switch 631 is connected in series between the first input terminal 611 and a node 64. The second switch 632 is connected in series between the second input terminal 612 and node 64. Node 64 is connected to the output terminal 62. The output terminal 62 is connected to the transmission line 3. The transmission line 3 is connected to the capacitive load 2.

[0096] The first and second switches 631 and 632 are special in that they are bidirectional. That is, when switches 631 and 632 are closed, they do not allow any current to flow, regardless of the direction of flow. Bidirectional switches are implemented, for example, using two MOSFET or HEMT transistors connected in a common-drain configuration or, preferably, in a common-source configuration.

[0097] Thus, if switches 631 and 632 are never closed simultaneously, the first and second terminals 631 and 632 are never connected to each other. Therefore, all the energy produced by generator 4 is transferred to the output terminal 62, but not to the voltage source. If the switches are unidirectional, meaning they only block current in one direction, they can transfer some of the energy injected from generator 4 to the voltage source, which can be a source of loss in circuit 1. Therefore, bidirectional switches 631 and 632 improve the control efficiency.

[0098] The first and second switches, 631 and 632, are controlled by the controller of circuit 1. For example, the controller is configured to close one of the switches while keeping the other open, and vice versa. The controller is preferably configured not to close both switches 631 and 632 simultaneously, in order to avoid the losses described above.

[0099] In this embodiment, the first voltage source 51 is configured to deliver a voltage that can take on different amplitudes, particularly as a function of time. The delivered voltage can have a first amplitude, for example 0 V, and a second amplitude, for example 6 V. The different voltage amplitudes correspond for example to the holding voltages, respectively "low" and "high", which it is necessary to apply to the grid of the power component to maintain the component in a blocked state or a conducting state.

[0100] In the embodiment of [Fig. 2], circuit 1 includes a first impedance 71 connected in series between node 64 and the output terminal 62 of the switch 6. The impedance serves to reduce the amplitude of a bounce originating from the transmission line. The bounce corresponds, for example, to a reflection of the current pulse on the component (i.e., on the capacitive load) because the component is not necessarily matched to the transmission line. In this way, the bounce is not sent back into line 3.

[0101] The first impedance 71 also protects the upstream elements (considering the side of the transmission line 3 as downstream of the impedance 71). Thus, in this embodiment, the first and second switches 631, 632 as well as the generators and sources at the inputs of the switch 6 are protected by the impedance 71.

[0102] The value of the first impedance depends on its placement in the control circuit and in particular on the impedance of the transmission line 3 which must be connected to the output terminal 62 of the switch 6. It is preferably equal to the impedance of the transmission line 3, for example typically between 10 Q and 100 Q. It is also preferably resistive.

[0103] Figure 3 presents a timing diagram of one operating mode of the control circuit 1. This figure shows in particular the synchronization of the different elements orchestrated by the controller during the injection of a current pulse to the transmission line 3. The timing diagram shows as a function of time the currents I4, 16 coming respectively from the pulse generator 4 and the switch 6, the voltage V51 across the terminals of the first voltage source 51 and the conducting state (represented by "ON") and blocked state (represented by "OFF") of the switches 631, 632 of the switch 6. The timing diagram also shows the voltage across the terminals of the capacitive load 2.

[0104] In the initial state, before the pulse generation is triggered, the first switch 631 is open (or blocked) and the second switch 632 is closed (or conducting). The voltage V51 produced by the first voltage source 51 is, for example, zero. This could be the component's low-holding voltage in a blocked state. The voltage V2 across the load 2 is also zero.

[0105] This initial state, where pulse generator 4 is disconnected from output 62, is conducive to the charging of generator 4. The controller has, for example, sent a charge command to pulse generator 4 to start its charging (area “charge” in figure 3).

[0106] At a given moment, the controller sends a pulse generation trigger command to the pulse generator 4; commands the first switch 631 to close (state "ON"); and commands the second switch 632 to open (state "OFF"). A high-amplitude current pulse Δ4 is ​​delivered by the pulse generator 4 for an initial duration te. Since the first switch 631 is conducting, the pulse Δ4 is ​​injected into the switch 6, which therefore delivers the current pulse le to the transmission line 3. Since the second switch 632 is open and is a bidirectional switch, the entire current Δ4 delivered by the pulse generator corresponds to the current le delivered by the switch 6. Therefore, there are no losses.

[0107] The first duration te can be between 10 ps and 100 ps, ​​typically between 1 ns and 40 ns. The amplitude of the current Î4 is, for example, equal to 2 A. However, it can be adjusted according to the load 2 to be controlled.

[0108] The first switch 631 remains closed for a second duration tp, which is, for example, longer than the first duration te. After the pulse is generated, the pulse generator 4 can apply a constant voltage to its output branch. Advantageously, the durations tp and te are equal. During the duration tp, the pulse generator 4 generates a small current Î4, which corresponds to the holding current on the load 2.

[0109] While the second switch 632 is open, the controller sends a command to the first voltage source so that it now delivers a voltage V51 with a second amplitude, greater than the first amplitude. The second amplitude can typically be between 3 V and 100 V, for example typically 3V or 15 V. This second amplitude corresponds for example to a high holding voltage of the power component switched in a conducting state.

[0110] At the end of the second duration tp, the controller commands the opening of the first switch 631 and the closing of the second switch 632. The switch 6 can deliver a low amplitude current le which corresponds to the load holding current 2). This holding current is for example from the first voltage source 51.

[0111] In the case where the voltage V2 reached by the grid 2 at the end of the pulse is not equal to the voltage V51 delivered by the first voltage source 51, a significant transient current can be established in order to complete the charging of the grid 2.

[0112] It is possible that the voltage V51 delivered by the first voltage source 51 will differ from the voltage delivered by the pulse generator 4 at the end of pulse generation. However, it is advantageous for these voltages to be sized so that the transition is continuous and differentiable. In this way, the risk of sending a spurious signal to the power component is reduced.

[0113] In response to this operation, the voltage V2 across load 2 increases sharply, following a steep edge, when the current pulse is injected into the capacitive load 2. The voltage then stabilizes when the pulse ends and a constant voltage is maintained. For comparison, Figure 3 also shows, in dashed form, the voltage V2 of the capacitive load if it were biased solely by the first voltage source 51. The charging is much slower than that obtained by the invention.

[0114] Opening the first switch 631 stops the pulse generation, even if it is still occurring. Thus, the second time tp (at the end of which the first switch 631 is open) imposes a maximum duration on the current pulse.

[0115] It is advantageous for the pulse generator 4 to be configured to maintain a constant voltage immediately after the current pulse generation. This voltage is preferably equal to the holding voltage, which is applied to the component. In this way, the component completes its switching or maintains its state while the first voltage source 51 takes over.

[0116] The first duration is measured from the generation of the current pulse. For example, it is measured from the sending of the trigger command by the controller of circuit 1.

[0117] The second time interval tp is measured from the transmission of the current pulse to output 62, that is, from the moment the first switch 631 closes. However, if the pulse generation is slightly delayed relative to the switch closure, the second time interval tp is measured from the moment the pulse is generated. In all cases, this time interval is measured from the later of the two events (closure or generation). Indeed, the injection of the current pulse to the output terminal 62 of the switch 6 only begins when both conditions (first switch closed and generation) are met.

[0118] Figure 4 schematically presents a third embodiment of the control circuit 1 according to the invention. Unlike the embodiment in Figure 2, the switch 6 also includes a third input terminal 613 which is connected to a second voltage source 52. The switch 6 also includes a third switch 633, connected between the third input terminal 613 and the output terminal 62, and more particularly to node 64 of the switch 6.

[0119] In this example, the first and second voltage sources 51 and 52 are configured to deliver a stable voltage. The first voltage source 51 is configured, for example, to deliver a voltage equal to the component's low-holding voltage in its off state. The second source 52 is configured, for example, to deliver a voltage equal to the component's high-holding voltage in its on state. Thus, instead of using a variable voltage source (as shown in [Fig. 3]), circuit 1 can implement fixed voltage sources.

[0120] In order to take advantage of both sources 51 and 52, the controller is also configured to control the third switch 633. For example, after sending the trigger command to the pulse generator 4, the controller can command the closing (by closing, we also mean holding closed) of the third switch 633 while keeping the first and second switches closed. Switches 631 and 632 are open. In this way, the voltage V52 supplied by the second source 52 is applied to the output terminal 62 of the switch 6 and, for example, into the transmission line 3. In order not to short-circuit the current supplied by the pulse generator 4, the controller can be configured to open the third switch 633 as soon as the trigger command is sent to the pulse generator 4 and then keep it open.

[0121] Advantageously, the first and second voltage sources 51, 52 are part of the control circuit 1.

[0122] In the embodiment of [Fig. 4], the circuit 1 includes first, second, and third impedances 71, 72, 73 that reduce a bounce from the output terminal 62 (and, for example, from the transmission line 3). Each impedance 71, 72, 73 is placed between one of the switches 631, 632, 633 of the switch 6 and its associated input terminal 611, 612, 613. In this way, the three impedances 71, 72, 73 are positioned on the three branches along which current waves can propagate. The proximity of the impedances to the pulse generators 4 and the voltage sources 51, 52 thus ensures a reduced distance to the ground points of the generators or sources. Impedances 71, 72, 73 preferentially have an impedance equal to the impedance of the transmission line 3. For example, they have an impedance between 10 Q and 100 Q. They are also preferentially resistive.

[0123] In an unrepresented development, the control circuit 1 may include a plurality of generators 4 connected to the switch 6. In this case, the switch 6 may include at least a fourth input and a fourth switch (assuming that the two voltage sources 51, 52 are retained). The fourth switch connects the fourth input to the output terminal 62. The fourth input is connected to the additional generator. The controller can be configured to operate only one of the two generators 4 at a time. For example, it allows the selection of one of the generators 4 to be implemented and then performs the switching of the load 2 considering only that generator.

[0124] The same principle applies to voltage sources. Circuit 1 can include several different voltage sources (for example, fixed or adjustable). These voltage sources are then all connected to the switch 6. The switch 6 advantageously includes a number of inputs and corresponding switches allowing these voltage sources to be connected to the output terminal 62. The controller can then be configured to allow the selection of the voltage source(s) to be used and to perform the load switching with this or these voltage sources.

[0125] Circuit 1, for example, as shown in its various embodiments in [Fig. 1], [Fig. 2], and [Fig. 4], can be implemented using discrete components and / or an integrated circuit, such as an application-specific integrated circuit (ASIC). Using an integrated circuit reduces the size and weight of circuit 1.

[0126] Figure 5 schematically presents an embodiment of a pulse generator 4 such that it can be implemented in circuit 1. The generator 4 comprises four switches 431, 432, 433, 434 and an inductor 45.

[0127] The inductor 45 is connected between two half-bridges formed by switches 431, 432, 433, and 434. The first half-bridge is formed by the first switch 431 and the second switch 432 connected in series. This half-bridge is intended to be biased between a first terminal 411 and a second terminal 412. The second half-bridge is formed by the third switch 433 and the fourth switch 434, also connected in series. This half-bridge is intended to be biased between a third terminal 413 and a fourth terminal 414.

[0128] The inductance 45 is connected between switches 431, 432 of the first half-bridge, at the level of a first node 421, and switches 433, 434 of the second half-bridge, at the level of a second node 422.

[0129] Inductance 45 can have a value of 150 nH, but its size and impedance can vary greatly depending on the intended applications.

[0130] Generator 4 also includes an output branch 46, connected to the second node 422 of the second half-bridge.

[0131] Generator 4 is notable in that switches 431, 432 of the first half-bridge are unidirectional and switches 433, 434 of the second half-bridge are bidirectional.

[0132] In particular, the switches 431, 432 of the first half-bridge each comprise a bidirectional switch 4311, 4321 and a diode 4312, 4322, referred to as a "body diode". Each of the bidirectional switches 4311, 4321 is connected in parallel with a body diode 4312, 4322. The body diodes 4312, 4322 are oriented so that they conduct current flowing in the same direction in the first half-bridge. In this case, in [Fig. 5], they conduct current flowing from the second terminal 412 to the first terminal 411.

[0133] Generator 4 also includes a controller, not shown in [Fig. 5]. This controller may also be an electronic circuit or a programmable computer as previously described, including communication means. The controller of generator 4 is configured to receive commands from the controller of control circuit 1. In one embodiment, the controller of control circuit 1 is also the controller of generator 4.

[0134] The generator controller 4 is further configured to control switches 431, 432, 433, 434 to charge inductor 45 and generate a current pulse by discharging the inductor on output branch 46. The controller can also be configured to maintain a constant potential on the output branch and / or stop pulse generation, even if inductor 45 is not fully discharged.

[0135] To make the most of the generator 4, each half-bridge can be biased by means of a potential difference applied respectively between the first and second terminals 411, 412 and the third and fourth terminals 413, 414. The generator 4 may include a power supply for performing these biases. Alternatively, the control circuit 1 may include a power supply for performing this task.

[0136] Regardless of the polarization method used, we will consider that the potentials V1, V2, V3 and V4 are respectively applied to the first, second, third and fourth terminals 411, 412, 413, 414.

[0137] The unique structure of this generator 4 (in particular the presence of two bidirectional switches) allows, notably, the generation of one or more arbitrary pulses. Indeed, the adjustment parameters of generator 4, for example the amplitude and width of the current pulse, as well as the levels of the high and low plateaus, are sufficiently numerous and varied to allow the generation of any pulse, therefore arbitrary. It should also be noted that this generator 4, considered alone, is capable of fulfilling a function similar to the control circuit 1 considered as a whole, which is itself arbitrary.

[0138] Given the orientation of the body diodes 4312, 4322 of the unidirectional switches 431, 432, it is preferable that V1 > V2. In this way, when the unidirectional switches 431, 432 are open, no current flows in the first half-bridge.

[0139] A potential difference V2 > V1 can cause a short circuit in the body diodes 4312, 4322 of the first and second unidirectional switches 431, 432. Therefore, it may be preferable for at least one of these two switches to be bidirectional. Thus, when the switch is open, there is no risk of a short circuit. When the first switch 431 is bidirectional, the operating mode of the generator 4 can be maintained to achieve positive charging and discharge of the inductor (see modes A to D in Figure 6). However, for certain operating modes, such as modes G and H in Figure 6, and in the absence of a body diode 4312, the first switch 431 must be specifically closed, for example by the controller, so that the inductor can discharge properly.

[0140] The two switches 431 and 432 can also be bidirectional switches. In this case, it is necessary to ensure that the controller closes the switches when they should be conducting in the absence of a body diode. For example, in operating modes C, D, G, and H in Figure 6, and in the absence of body diodes 4312 and 4322, the first and second switches 431, 432 must be specifically closed by the controller so that the inductance can discharge properly.

[0141] The half-bridge comprising the bidirectional switches 433 and 434 can be polarized in either direction. For this example, we will consider V3 > V4.

[0142] In order to perform the charging or discharging of the inductor, it is preferable that the half-bridges be biased so that V1 > V4 and V3 > V2. For example: V2 = V4 = 0 V and V1 = V3 = 200 V.

[0143] In the embodiment of [Fig. 5], the generator 4 also includes an impedance 70 on the output branch 46 whose role is to reduce bounces from the circuit 1, such as a bounce from a transmission line 3. Similar to the first, second and third impedances 71, 72, 73 described previously, the impedance 70 is arranged on a branch connecting two fixed potentials.

[0144] Figure 6 schematically presents, with reference to the letters A to H, different operating modes of generator 4 depending on the configuration of switches 431, 432, 433, and 434. Reference symbols are shown only on one of the subfigures for clarity. However, each element of the generator (except for the impedance 70) is also shown in Figure 5. In these operating modes, the first and second switches 431 and 432 are unidirectional. The potentials applied to terminals 411, 412, 413, and 414 satisfy the biases V1 > V2 and V3 > V4. The bold line represents the flow of an electric current in generator 4. An arrow indicates the direction of the current (considering the aforementioned polarization convention).

[0145] The first phase of operation corresponds to operating modes A to D.

[0146] Operating mode A corresponds to the charging of the inductance 45. The controller commands the closing of the first and fourth switches 431, 434. A current flows through the inductance 45 and the inductance 45 stores a quantity of energy.

[0147] The coil can be energized by the generator receiving an energizing command. This energizing command... charge is preferentially sent by the controller of the control circuit 1. With reference to [Fig. 3], this command is sent in the charging period, before the triggering of the current pulse generation.

[0148] Operating mode B corresponds to the injection of a current pulse into the output branch 46 by the initially charged inductor 45. This pulse is triggered by the opening of the fourth switch 434. The inductor 45 discharges into the load connected (via the switch 6) to the output branch 46. For example, the transmission line 3 and the capacitive load 2.

[0149] The fourth switch 434 opens upon receiving a trigger command for pulse generation. Referring to [Fig. 3], this command is sent by the control circuit 1 to the generator controller 4. This command transmission coincides with the change of state of the switch 6, so that the current pulse is sent to the load 2.

[0150] In operating mode C, the inductor 45 also discharges into the output branch 46. This operating mode is triggered when the inductor 45 is charged, or sufficiently charged, and the first and fourth switches 431 and 434 are open. For example, the first switch 431 is opened by a control signal. The inductor 45 continues to discharge, delivering current into the only accessible branch, that is, through the body diode 4322 of the second switch 432. In this operating mode, the second switch 432 is conducting, even though its internal switch 4321 is blocked.It is noted that the opening of the first switch 431 to trigger the discharge of the inductor 435 can be carried out at any time between the opening of the fourth switch 434, carried out during the implementation of operating mode B, and before the closing of the third switch 433, carried out during the implementation of operating mode D described later.

[0151] This operating mode can be the operating mode used to generate the current pulse on the output branch 46. It allows the inductance 45 to discharge, at least partially, into the output branch 46.

[0152] This operating mode, however, does not allow for a new load on inductor 45 to achieve a reverse switching of load 2 (if the first switching was, for example, from blocked to conducting, the second switching is from conducting to blocked). Operating mode C is therefore preferentially a transient operating mode between operating modes B and D.

[0153] The energy stored in inductor 45 can depend on the charging time of inductor 45. Thus, by controlling the charging time of inductor 45, the amplitude of the generated current pulse can be controlled. For certain applications, it may be advantageous to avoid sending a pulse with too high an amplitude. Indeed, it is worth remembering that some power components, such as HEMT transistors, can be damaged by excessively high voltage or current spikes.

[0154] The charging time can be controlled by a duration between the receipt of a charging command and the trigger command for pulse generation.

[0155] Alternatively, the pulse generator 4 includes a first current limiter (not shown in the figures). This first limiter is connected in series with the inductor 45 so as to limit the current flowing through the inductor 45 when it is under load. In operating mode A, corresponding to the load, the limiter is, for example, inserted between the first and fourth terminals 411, 414. If it is desired that the current pulse not be limited during generation but only during the load, it is preferable for the limiter to be inserted in the second half-bridge, between the second node 422 and the fourth terminal 414.

[0156] To limit the charging current regardless of the current's sign, it is preferable for generator 4 to include a second current limiter. This could also be inserted in the second half-bridge, between the third terminal 413 and the second node 422.

[0157] If it is desired that the limiter limit the amplitude of the generated pulse rather than the amplitude of the load current, then it is preferable to insert the first limiter between the two half-bridges or between the first terminal 411 and the first node 421. Alternatively, when generator 4 includes two limiters, each of them can be inserted in the first half-bridge between, respectively, the first node 421 and the first terminal 411 and between the first node 421 and the second terminal 412.

[0158] In operating mode D, the third switch 433 is closed, for example, by a closing command for the third switch 433 issued by the controller. The inductor continues to discharge, delivering a current flowing from V2 to V3. As soon as the current becomes zero, the body diode 4322 of the second switch 432 prevents a reverse current (which would naturally flow from V3 to V2) from forming, and the inductor 45 remains discharged.

[0159] Closing the third switch 433 imposes the potential V3 on the output branch 46. This potential setting allows the establishment of a holding current (if V3 > V4 it is the high holding voltage) in the load 2 so that the latter maintains its state.

[0160] Unlike mode C, mode D does not wait for the complete or substantial discharge of inductor 45 before applying a holding voltage to the component. It is applied at an arbitrary instant, as soon as the third switch 433 is closed. Closing the third switch 433 stops the current pulse from being injected into the branch and fixes the potential of the output branch 46.

[0161] Closing the third switch 433 acts as a stop to the pulse injection in the circuit. The control circuit 1 can, for example, send a command to stop pulse generation to the generator controller 4 so that the latter commands the third switch 433 to close.

[0162] Note that a holding current, which may be weak and constant, may continue to flow in the output branch 46.

[0163] The charging acceleration of inductance 45 can be achieved by using a high potential difference V1 - V4. This reduces the charging time and allows the switching frequency of the components to be increased.

[0164] A second phase, corresponding to modes E to H, allows for a new charge of the inductor 45, the injection of a current pulse onto the output branch, and the completion of the inductor's discharge and / or the termination of the pulse injection. However, in this second phase, the direction of the current flowing in The inductance (and therefore the current injected into the output branch) is reversed with respect to the direction of the current in mode A. This is referred to as a negative charge. This is achieved, in particular, by charging the inductance 45 between potentials V2 and V3, where V2 < V3.

[0165] It is noted that operating modes E to H are implemented similarly to operating modes A to D, but with a different (in this case, symmetrical) switching of the four generator switches. Specifically: Operating mode E, comparable to operating mode A, is implemented by closing the second and third switches 432 and 433 and by opening the first and fourth switches 431 and 434; the opening and closing of these switches are, for example, carried out, as for operating mode A, by one or more commands issued by the controller; Operating mode F, comparable to operating mode B, is implemented by opening the third switch 433; this opening is, for example, carried out, as for operating mode B, by a command issued by the controller; Operating mode G, comparable to operating mode C, is implemented by opening the second switch 432; this opening is, for example, carried out, as for operating mode C, by a command issued by the controller; Operating mode H, comparable to operating mode D, is implemented by closing the fourth switch 434; this closure is, for example, carried out, as for operating mode D, by a command issued by the controller;

[0166] The generator controller 4 is therefore configured to: upon receiving a load command, charge the inductor 45 according to operating mode A (or E respectively); upon receiving a trigger command to generate a current pulse, discharge the inductor 45 into the output branch 46 according to operating mode B or C (or respectively F or G); and if there is Heu, upon receiving a command to stop the generation of the pulse, stop the generation of the pulse according to operating mode D (or respectively H).

[0167] In addition, the controller can also be configured to, upon receiving a new charging command: re-charge the inductance 45 positively (or respectively negatively) in the same way, according to operating mode A (or respectively E); or negatively (or respectively positively) charge the inductance 45 with a current flowing in the opposite direction, according to operating mode E (or respectively A).

[0168] According to one variant, the first switch 431 (or the second switch 432) is opened after the generation of the positive pulse in phase B (or after the generation of the negative pulse in phase F). Thus, the discharge phase C (or G) is not required to complete the discharge of the inductor, and the generator 4 can proceed directly to phase D (or H). This variant exhibits energy performance equivalent to the implementation shown in [Fig. 6] and can allow for an increased pulse generation frequency.

[0169] Generator 4 may include a circuit 1 implemented using an integrated circuit. Similarly, generator 4 may be implemented using discrete components and / or an integrated circuit, such as a specialized integrated circuit. Using an integrated circuit reduces the size and weight of generator 4.

Claims

CLAIMS

1. Pulse generator (4) of current on an output branch (46) upon receipt of a trigger command, comprising: first, second, third and fourth terminals (411, 412, 413, 414) configured to be connected to, respectively, first, second, third and fourth electrical potentials V1, V2, V3, V4; first, second, third and fourth switches (431, 432, 433, 434); an inductor (45); the output branch (46);and a controller, the first switch (431) being connected between the first terminal (411) and a first node (421), the second switch (432) being connected between the second terminal (412) and the first node (421), the third switch (433) being connected between the third terminal (413) and a second node (422), the fourth switch (434) being connected between the fourth terminal (414) and the second node (422), the inductor (45) being connected between the first node (421) and the second node (422), the output branch (46) being connected to the second node (422), the pulse generator (4) being characterized in that the third and fourth switches (433, 434) are bidirectional switches and in that the controller is also configured to: charge the inductor by closing the first and fourth switches (431, 434) and by opening the second and third switches (432, 433);and upon receiving the trigger command, discharge the inductance in the output branch by opening the fourth switch (434).

2. Pulse generator (4) according to claim 1, wherein the controller is also configured to, after the opening of the fourth switch (434), close the third switch (433) and open the first switch (431).

3. Pulse generator (4) according to any one of the preceding claims, wherein the controller is also configured to: negatively charge the inductance by closing the second and third switches (432, 433) and opening the first and fourth switches (431, 434); and discharge the inductance again into the output branch by opening the third switch (433).

4. Pulse generator (4) according to the preceding claim, wherein the controller is also configured to: after opening the third switch (433), open the second switch (432) and close the fourth switch (434).

5. Pulse generator (4) according to any one of the preceding claims, wherein the third and fourth switches (433, 434) each comprise two transistors mounted in common sources.

6. Pulse generator (4) according to any one of the preceding claims, comprising an impedance (70) for reducing the amplitude of a bounce from the output branch (46).

7. Pulse generator (4) according to any one of the preceding claims, comprising a first current limiter connected in series with the inductor (45) when the first, second, third and fourth switches (431, 432, 433, 434) are in a configuration (A; E) permitting the inductor (45) to be loaded.

8. Pulse generator (4) according to any one of the preceding claims, comprising a power supply configured to apply the first, second, third and fourth electrical potentials V1, V2, V3, V4 on, respectively, the first, second, third and fourth terminals (411, 412, 413, 414), with V1 > V4 and V3 > V2.

9. Control circuit (1) of a power electronic component through a transmission line (3), the control circuit (1) being characterized in that it comprises: a pulse generator (4) according to any one of the preceding claims; a switch (6); and a controller, the switch (6) comprising: an output terminal (62) configured to be connected to the transmission line (3); a first input terminal (611) and a second input terminal (612), the first input terminal (611) of the switch (6) being configured to be connected to the output branch (46) of the pulse generator (4), the second input terminal (612) of the switch (6) being configured to be connected to a first voltage source (51); and a first bidirectional switch (631) and a second bidirectional switch (632), the first switch (631) being connected between the first input terminal (611) and the output terminal (62), the second switch (632) being connected between the second input terminal (612) and the output terminal (62), the controller being configured to: send a trigger command to the pulse generator; upon sending the trigger command to the pulse generator, close the first switch (631) and open the second switch (632); and an instant after sending the trigger command to the pulse generator (4) or after closing the first switch (631), open the first switch (631) and close the second switch (632).

10. Control circuit (1) according to the preceding claim, comprising a first impedance (71) for reducing the amplitude of a bounce from the transmission line (3) on the first input terminal (611) and the second input terminal (612). [Claim 1 1] Control circuit (1 ) according to claim 10, wherein the first impedance (71 ) is connected between the first and second switches (631 , 632) and the output terminal (62) of the switch (6).

12. Control circuit (1) according to claim 10, also comprising a second impedance (72), the first impedance (71) being connected between the output terminal (62) and the first input terminal (611), preferably between the first input terminal (611) and the first switch (631), the second impedance (72) being connected between the output terminal (62) and the second input terminal (612), preferably between the second input terminal (612) and the second switch (632).

13. Control circuit (1) according to any one of claims 9 to 12, wherein the controller is configured to vary the amplitude of the voltage delivered by the first voltage source (51) as a function of time, between a first amplitude and a second amplitude different from the first amplitude, the controller being further configured so that the variation between the first amplitude and the second amplitude is achieved when the second switch (632) is open.

14. Control circuit (1) according to the preceding claim, wherein the controller is also configured to close the second switch (632) before the generation of the current pulse.

15. Control circuit (1) according to any one of claims 9 to 12, wherein the switch (6) further comprises: a third input terminal (613) configured to be connected to a second voltage source (52); and a third bidirectional switch (633), connected between the third input terminal (613) and the output terminal (62), the controller also being configured to: before sending the trip command to the pulse generator, close the second switch (632) and open the first and third switches (631, 633); upon sending the trip command to the pulse generator, close the first switch (631) and open the second switch (632); and an instant after sending the trip command to the pulse generator, close the third switch (633) and open the first and second switches (631, 633).

Images