How to protect current-fed converters from overvoltages

Abstract

The present disclosure relates to a method for protecting a current-fed converter comprising a switching cell having two or more legs, each leg being provided with at least one switching device, from overvoltages, the method comprising: detecting a current differential signal dI in each of the upper or lower switching devices of at least two of the legs in which a switching transition occurs, in an open circuit fault condition of a switching device in a leg of the switching cell; Q / dt, where the switching transition includes a change in a current conducting switching device in an upper switching device or a lower switching device, and triggering protection during the switching transition based on the current differential signals when either an absolute value of one of the current differential signals is smaller than a first predetermined value k1 or is null, or a sum of the absolute values ​​of the current differential signals is smaller than a second predetermined value k2. The present disclosure also relates to an electronic circuit configured to perform the method.

Description

[Technical field]

[0001] The present invention relates to high power current supply converters and to overvoltage protection for such converters. [Background technology]

[0002] Current-fed converters are widely used in various applications such as HVDC, motor drives or power factor correctors. The main limitation of such converters was the use of large input inductances. However, the emergence of wide bandgap semiconductor devices, e.g. silicon carbide and / or gallium nitride semiconductor devices, with lower conduction and switching losses, has opened new perspectives in the design by miniaturizing and improving the efficiency of current-fed converters. Nevertheless, the use of insulated gate bipolar transistors (IGBTs) is still state of the art for high power applications, where there is still room for improvement, especially in protection measures.

[0003] In a current-fed power converter consisting of at least two legs, the switches of the different legs are controlled from conducting to blocking and from blocking to conducting in an independent and complementary manner. An overlap (both switches ON) is used during the switching transitions (or switching events) between the two switches. The main reason for the overlap is due to the inductive nature of the input source. A DC link current interruption can be due to multiple factors including a bad control signal of one of the two switches, a faulty device failing in an open circuit, or a gate driver failing. As a result, there is no current path for the DC current, which can lead to severe damage. A current interruption in a current-fed converter induces a rapid rise in DC voltage higher than the breakdown voltage of the power devices, causing the destruction of the power converter, because they cannot withstand such high voltage and high current conditions for such a long time to enter the avalanche mode and dissipate the energy stored in the inductance. A current interruption in a current-fed converter is an occasional occurrence and is a fault that most often leads to the breakdown of the converter.

[0004] In addition, the increasing use of high-speed IGBTs (Integrated Gate Commutated Thyristors), which are faster than their corresponding IGCTs, in high-power applications requires faster response times of protection functions. To avoid DC voltage rise and destruction of the power converter during current interruption, high-speed protection circuits must be implemented that ensure a freewheeling path for the DC link current. Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, there is a need for a fast detection protection circuit, and the present invention addresses this need.

[0006] The problem is that typical protection circuits are large and require several power devices such as diodes, TVS (transient voltage suppressors), thyristors, and metal oxide varistors (MOVs).

[0007] In most cases, such circuits are only used to clamp overvoltages within a tolerance range, but do not adequately detect faults that would shut down the converter. Moreover, these protection circuits rely on voltage measurements, which means that an overvoltage event is already in progress.

[0008] Furthermore, some of the conventional solutions introduce parasitics due to the large number of power components (diodes, etc.), which increases losses since the addition of parasitic capacitors increases the switching losses of the converter. In some cases, protection circuits also conduct part of the source current, which may be the case for current source inverters.

[0009] In addition, solutions based on detecting and ensuring a freewheeling path for the DC link current by the converter's controller are not fast enough.

[0010] The objective of the present disclosure is to quickly ensure overvoltage protection due to current interruption in current-fed converters. The approach to solving the problem aims to prevent as many voltage rises as possible by monitoring the root cause, which is the current interruption.

[0011] The present disclosure also aims to limit the use of large power devices such as diodes, thyristors or TVS by using low voltage analog and logic circuits for detection that control only one power device that ensures a freewheeling path for the current. [Means for solving the problem]

[0012] More precisely, the present disclosure relates to a method for protecting against overvoltages a current-fed converter comprising a switching cell having two or more legs, each leg being provided with at least one switching device, the method comprising the steps of: A current differential signal dI in each of at least two upper or lower switching devices of the leg in which a switching transition occurs. Q / dt, where a switching transition includes a change in current conducting switching device in an upper switching device or a lower switching device; During the switching transition, based on the current derivative signal, When the absolute value of one of the current differential signals is less than a first predetermined value k1 or is null, or When the sum of the absolute values ​​of the current differential signals is smaller than a second predetermined value k2, Triggering protection when either The present invention relates to a method comprising the steps of:

[0013] Such a method can be easily implemented using analog and logic components for fast detection and triggering of protection.

[0014] The method may include synchronizing measuring the current derivative signal with switching transitions between switching devices generated by a controller of the converter.

[0015] Synchronizing includes adjusting a blanking window that starts with the onset of a switching transition and ends before, at or after the end of the switching transition, which blanking window inhibits triggering of protection during the switching transition.

[0016] The method may include measuring the sign of the output voltage across a switching cell to identify a leading device and a freewheeling device in the switching cell and define the width of the blanking window.

[0017] Triggering the protection may include providing a current path external to the switching cell for a current source supplying current to the converter.

[0018] The method may include recording the fault condition in the converter controller after triggering the protection and / or forwarding the fault condition of the converter in the remote controller.

[0019] The present disclosure also provides an electronic device for carrying out the above-mentioned method, A means for measuring a current differential signal; an electronic circuit configured to compare the current derivative signal with a first predetermined value or a second predetermined value to trigger a protection; a protection device for providing an alternative path for the current provided by the current source outside the switching cell when the protection is triggered by the electronic circuit; The present invention relates to an electronic device comprising:

[0020] In the above electronic device, a. the means for measuring the current derivative signal may comprise a sensor for sensing a current derivative of a current flowing through each leg of the switching cell, and means for adjusting the current derivative signal issued from the sensor; b.The electronic circuit is means for providing a first absolute value of the first current derivative signal and means for providing a second absolute value from the second derivative signal; a first comparator for comparing the first absolute value with a first predetermined value K1 and providing a first logic output in response to an output state of the first comparator; a second comparator for comparing the second absolute value with the first predetermined value K1 and providing a second logic output in response to an output state of the second comparator; and a logic gate connected to the outputs of the first comparator and the second comparator for providing a trigger signal for triggering a protection device when either one of the absolute values ​​is smaller than the first predetermined value. Or, summing means for providing the sum of the absolute values; and a third comparator for comparing the sum with a second predetermined value and providing a third logic output in response to an output state of the third comparator, the third comparator providing a trigger signal for triggering a protection device when the sum of the absolute values ​​is less than the second predetermined value. Either It can be provided with:

[0021] The sensor that detects the current derivative is a Rogowski coil.

[0022] The means for conditioning the current derivative signal may comprise a filter and / or an amplifier.

[0023] The means for measuring the current differential signal and the electronic circuit are preferably constructed of analog circuits and logic circuits, which allows high-speed detection.

[0024] The protection device may comprise a protection switch that freewheels the current supplying to the switching cell.

[0025] The electronic device may comprise a gate for enabling the trigger signal, the gate having the trigger signal as a first input and a blanking signal as a second input for suppressing the trigger signal during at least a portion of a switching transition issued from the blanking circuit, the blanking signal being issued from the converter controller, the gate being configured to suppress the trigger signal during occurrence of the blanking signal.

[0026] The present disclosure also relates to a current-fed converter comprising an electronic device, the electronic device being provided on a PCB substrate within the converter.

[0027] The proposed method can be applied to any current-fed converter consisting of at least two legs, each leg containing at least one controllable switching device, which are alternatively controlled and include during normal operation an overlap of control signals that ensures current continuity in the switching device where the transition occurs.

[0028] Multiple combinations of the current differential sensor output can be implemented to design overvoltage protection. The logic and conditioning circuits must be designed according to the signal combinations.

[0029] A detailed description of exemplary embodiments of the present invention is discussed below with reference to the accompanying drawings. [Brief description of the drawings]

[0030] [Figure 1] FIG. 1 shows a simplified two-leg current-fed converter design. [Diagram 2] 1 is a simplified diagram of the top of a current-fed converter having a current differential detection sensor; [Figure 3A] FIG. 1 illustrates a first embodiment of method steps according to the present disclosure. [Figure 3B] FIG. 2 illustrates a second embodiment of method steps according to the present disclosure. [Figure 4] FIG. 2 shows more detailed method steps in the first embodiment. [Diagram 5] FIG. 13 shows waveforms due to a first transition event without an open circuit fault. [Figure 6] FIG. 6 illustrates the first transition event of FIG. 5 with an open circuit fault. [Figure 7] FIG. 13 shows waveforms due to a second transition event without an open circuit fault. [Figure 8] FIG. 8 illustrates the second transition event of FIG. 7 with an open circuit fault. [Figure 9A] FIG. 2 illustrates one embodiment of a fault detection circuit. [Figure 9B] FIG. 2 illustrates one embodiment of a fault detection circuit. [Figure 10] FIG. 1 illustrates a Rogowski coil applicable to the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] The present invention relates to a method and device for preventing overvoltages in current-fed converters.

[0032] A simplified schematic diagram of such a converter is shown in Figure 1. In Figure 1, the converter comprises a switching cell fed by a DC voltage source 1 through an inductance 2 and having two legs. Each leg comprises an upper branch with a respective switching device 31, 32 and a lower branch with a respective switching device 41, 42. These switching devices are IGBTs or the like. Each branch is further provided with a series diode 71, 72, 81, 82 to block reverse voltages appearing across the IGBTs.

[0033] The gates of the IGBTs are under the control of gate drivers 51, 52 for the upper IGBTs 31, 32 and gate drivers 61, 62 for the lower IGBTs 41, 42. Each gate driver is under the control of a controller which provides gate driver control signals in accordance with the switching sequence of the converter to provide alternating current to the load 9.

[0034] As can be seen in the magnified view window 11, if bidirectional current conduction is required, the IGBT and series diode combination can be replaced by two common emitter IGBTs 32'a, 32'b and an anti-parallel diode 72'. The following description is based on an IGBT application but can be extended to any unipolar JFET, IGFET, HEMT, MOSFET 32"a, 32"b or bipolar BJT transistor, only the electrode names need to be changed.

[0035] Figure 2 is a simplified schematic diagram of the upper part of a switching cell with upper switch devices Q1 31 and Q3 32 provided with gate drivers 51, 52 and current differential sensing means 91, 92 according to the present disclosure arranged in series with them. The current source including voltage source 1 and inductance 2 of Figure 1 is indicated with reference 1' and the protection device is indicated as protection switch 93. A similar circuit is provided in the lower part of the switching cell with Q2 and Q4. The following method is applied in the same way both to the upper part of the cell and to the lower part of the cell. Figures 3A and 3B show the basic principle of the method of the present disclosure when there is a switching event between the switches Q1, Q3 of the upper part of the leg of the switching cell.

[0036] 3A discloses a first embodiment. In the first embodiment, the method includes: detecting a switching event between Q1 and Q3 in step 100; measuring current differential signals at Q1 and Q3 in step 110; comparing the absolute value of such current differential signal with a first preset value k1 in step 120a; and triggering protection in step 130 if one of the absolute values ​​of the two differential signals is smaller than the first preset value k1.

[0037] 3B discloses a second embodiment of the method, in which the sum of the absolute values ​​of the two current differential signals is compared with a second preset value k2, and if such sum is smaller than the second preset value k2, protection is triggered in step 130.

[0038] Although the method is shown as a flow chart, the measuring step 110, comparing steps 120a, 120b and triggering step 130 are performed using analog and logic components rather than software embedded in the converter's controller in order to provide real-time response.

[0039] Step 100 is a step initiated by the controller 10 of the converter to change the conduction state between the two switch devices 31 and 32 of the switching cell. The first switching event includes Q1 changing from a blocking state to a conducting state and Q3 changing from a conducting state to a blocking state with a delay relative to Q1, and the second switching event includes Q3 changing from a blocking state to a conducting state and Q1 changing from a conducting state to a blocking state with a delay relative to Q3.

[0040] Adding the two absolute values ​​has the advantage of providing a higher signal / noise ratio and a simpler detection circuit.

[0041] FIG. 4 provides an example of a first embodiment with more detail and includes provisions for setting a blanking window in step 104 to inhibit fault signals from triggering protection when a switching event has not yet ended, as can be seen in FIGS. 5-8 showing signal waveforms.

[0042] Such a blanking window is determined in advance in step 102 by the output voltage V OUT The duration depends on the polarity of the signal.

[0043] As soon as the blanking window is set, the detection signal V COMPis reset by the controller 10 in step 106 and is ready for detection. This detection signal is then set or not set depending on the presence of a sufficiently high dI / dt signal in the monitored leg of the converter. It should be noted that the steps of resetting the detection signal and setting the blanking window are provided in the controller of the converter. The polarity check can be performed before the start of the switching event and the blanking window is set by the controller at or before the start of the switching event. After setting the blanking window, if the absolute values ​​of both dI / dt are detected above a predefined value k1 in step 120a, V COMP The signal is set in step 122 and step 124 is performed after the blanking window ends. In such a case, protection is not initiated and the method includes waiting for another switching event. If the dI / dt signal is too small and V COMP If it is not possible to set such a detection signal will remain in a reset state and converter protection will be triggered in step 130.

[0044] FIG. 9A provides a schematic diagram of a first embodiment of a detection circuit of the present disclosure.

[0045] Starting from the two leg sensors 91, 92, the circuit comprises an amplifier circuit 12a, 12b, an absolute value detector or rectifier 13a, 13b and a latched comparator 14a, 14b. The comparators 14a, 14b have a reset signal input for a reset signal 19 from the controller of the converter and an input of a first predefined signal value k1 between which the two absolute values ​​are compared. The logic output of the latched comparator is input to a logic AND gate 15. Its output 13 is input to an OR gate 16. The OR gate 16 has as a second input the output signal 20 of the blanking circuit 18 of the controller 10. The output of the OR gate 16 provides a final trigger signal that triggers a protection device 93 if one of the two current differential signals is too weak.

[0046] In this example, the output signal of AND gate 15 provides a logic 0 trigger signal, while the blanking signal is a logic 1 signal which makes gate 16 an OR gate that suppresses the trigger signal until the blanking window ends. Other logic gates may be used in the case of a logic 1 trigger signal and / or a logic 0 blanking signal.

[0047] 9B corresponds to a second embodiment, in which the outputs of the absolute value detectors or rectifiers 13a, 13b are added in an adder 22. The result is compared in a latched comparator 14c. In such a case, the logic output of the comparator 14c becomes the input of an OR gate 16 together with a blanking window signal 20.

[0048] In the embodiment of Figures 9A, 9B, absolute value functions 13a, 13b are used to obtain a positive signal from the sensor, regardless of the sign of the current derivative. Several methods are possible to implement such a function, such as a full-bridge diode rectifier or a precision operational amplifier rectifier. Any circuit that can obtain the absolute value of any AC signal can be used.

[0049] The latching comparators 14a, 14b, 14c are similar to sample and hold circuits and can be implemented by using any kind of peak detector circuit or any comparator with a latch pin that keeps the output in a high state. Since the signal from the current differential sensor can be higher than the maximum rating, a clamp circuit can be inserted at the input for safety reasons.

[0050] The summation 22 of the current differential sensor outputs can be accomplished by connecting the sensors in series or by using a dedicated circuit using operational amplifier circuitry.

[0051] Regarding the sign module 17 and the blanking module 18, in most cases current-fed converters are controlled by vector modulation in the controller 10. This means that the switching devices are known by the controller, even in multiphase converters, which allows the sign module and the blanking module to be handled by the controller. In addition, the sign of the output voltage, and in particular the sign of the voltage across the switching cell, is obtained, which allows the leading and freewheeling devices in the switching cell to be identified. The duration of the blanking window can therefore be adjusted in the controller to be able to protect the converter in all switching configurations, as will be seen below.

[0052] The AND function 15 is needed with the first embodiment because one comparator is used per dI / dt sensor. If at least one of the two outputs of the comparator is low, this output triggers the protection.

[0053] The OR function 16 is used together with the blanking signal to avoid false triggering. This function can be implemented using simple diodes, dedicated ICs or discrete transistors.

[0054] The reset signal RST 19 is used to pull down the comparator output before initiating a new switching event commutation.

[0055] The protection device 93 is used as a freewheeling path for the DC current to avoid voltage overshoots. This device can be of the same type as the power device (IGBT, MOSFET, etc.) used in the converter.

[0056] The amplifiers 12a, 12b may be equipped with low pass filters to prevent transient signals from producing false detections.

[0057] The operation of the method on a signal is disclosed in Figures 5 to 8 for two different types of switching events: switching event ST1 in Figures 5 and 6 and switching event ST2 in Figures 7 and 8. Figures 5 and 7 represent switching events without an open circuit fault, while Figures 6 and 8 show switching events where a fault occurs. The description is made according to the following assumptions and initial conditions: An IGBT device with a series diode is considered to be in the switching cell; Load 9 V OUT The output corresponding to is taken as constant on the switching time scale. The sign of the output voltage is known. · The DC current is considered constant during the transition time scale. · Switch Q1 is considered a leading device which means that current variations occur only due to changes in Q1. · The protection device is represented as an open switch in normal operation and can be one or more controllable semiconductor devices.

[0058] The following discussion considers that the absolute current derivatives are summed and compared to a predetermined value k2.

[0059] In the two types of switching events or transitions discussed, Q1 is considered as the leading device and Q3 is considered as the freewheeling device. A similar explanation applies to the reverse situation. Figures 5 and 6 show switching transitions ST1 with turn-on of Q1 under normal and fault conditions, respectively. Figures 7 and 8 show switching transitions ST2 with turn-off of Q1 under normal and fault conditions, respectively. The detection / protection circuitry operates differently according to the switching transition. Indeed, the knowledge of the sign of the output voltage allows the leading device and therefore the switching type to be identified.

[0060] In Figure 5, during normal operation when 0 ≦ t < t1, the upper switch Q3 is ON and the switch Q1 is OFF. The gate voltages V GE3 210 and V GE1 200 are in the high (positive) state and the low (negative) state, respectively. The DC current flows from the collector (C) to the emitter (E) through Q3, which means that I Q3 240 is equal to I DC . At t = t1, a control signal from the controller is sent to Q1, which corresponds to the overlap point. At t = t1, the blanking window signal V BLANK 220 also toggles from the low state to the high state, instantaneously disabling the overvoltage detection circuit. After a certain delay (turn-on delay time or turn-off delay time) depending on the power device, the current 250 flowing through Q1 increases with a positive current differential, while the current differential of Q3 is negative in normal operation. The dI(t) / dt sensor arranged as shown in Figure 2 reacts, and as a result, the sum of the absolute values of the sensor output voltage increases as represented by the signal 270. In normal operation, the sum of the absolute values of the detected dI / dt is greater than the fixed reference k2, and thus the comparator output V COMP 230 is latched until the next commutation. At t = t2, the gate voltage of Q3 is pulled down, which means that the overlap time has ended, and the blanking window V BLANK 220 that progresses from t1 to t2 in the case of such a positive transition of Q1 also ends.

[0061] In the fault state of Q1 as shown in Figure 6, the switch Q1 cannot turn on. The initial conditions are the same as above, and the current 241 of Q3 is the DC current of the load. The blanking window V BLANK 220 starts at t1. Contrary to normal operation, no change occurs in the Q1 current 251 between t1 and t2, and V GE1 201 remains low. As a result, abs(dI Q1 / dt) is zero. The comparator output V COMP231 is not latched and remains reset. At t=t2, the gate signal 211 of Q3 is pulled down as a blanking signal, meaning that a fault has occurred. The protection device is triggered before the DC current is about to be interrupted.

[0062] In Figure 6, the commutation of Q3 should have occurred at t3, but abs(dI Q3 Even if k2 / dt) 271 is greater than the reference k2, the converter is shut down because the protection is turned on at t2. The current transition between t3 and t4 can be said to correspond to the current flowing from Q3 to the freewheeling circuit provided by the protection circuit.

[0063] The turn-off sequences shown in Figures 7 and 8 are slightly different. In Figure 7 without a fault, before t1, switch Q1 is ON and switch Q3 is OFF. The gate voltage V GE1 102 and Q3 gate voltage V GE3 112 are in a high (positive) state and a low (negative) state, respectively. DC current flows through Q1 from the collector (C) to the emitter (E), which is I Q1 =I DC At t=t1, a control signal from the controller is sent to Q3, which corresponds to the overlap instant, and both switches Q1 and Q3 are able to pass DC current. Despite the state change of Q3, nothing happens and current continues to flow through Q1. At t=t1, the blanking signal V BLANK122 also toggles from a low state to a high state, momentarily disabling the overvoltage detection circuit. At t=t2, the gate voltage of Q1 is pulled down, which means that the overlap time is over. After some delay (turn-on delay time or turn-off delay time) depending on the power device, at t=t3, the current flowing through Q3 increases with a positive current derivative, while the current derivative of Q1 is negative in normal operation. The dI(t) / dt sensor arranged as shown in FIG. 2 reacts, so that the sensor output voltage increases as represented by the dI / dt signal 272 higher than the predefined voltage value k2, thus increasing the comparator output V until the next commutation. COMP This triggers latching 232.

[0064] In the case of such a negative transition of Q1, the current transition occurs between t3 and t4, but this transition is initiated by the controller trying to turn on Q3, so the blanking window has duration from t1 to t4.

[0065] Switch Q3 does not turn on the DC current and V GE3 In abnormal operation such as in FIG. 8 where V213 remains low, Q1 turns off differently. In fact, the semiconductor GE1 203 tries to interrupt the current at t2, but the inductive nature of the source forces the current. The potential at the collector of Q1 rises, forcing Q1 into avalanche mode. As a result, the output signals from the current differential sensors to Q1 and Q3, respectively, are lower than normal and pseudo-null, with dI / dt273 being lower than the predefined voltage value k2. Therefore, the comparator output cannot latch before the end of the blanking signal, which would cause the circuit to trigger the protection device.

[0066] According to these analyses, the fault occurs when V COMP and V BLANK It can be said that this always occurs after the last falling edge of the control signal issued at the output of the AND gate having

[0067] The typical duration of an IGBT switching transition event is: t1~t2: 10μs~several hundred microseconds, overlap time; t2~t3: depends on the device, rated voltage, rated current, chip design and gate driver performance, corresponding to the turn ON / OFF delay time in the range of 10μs~100μs; t3 to t4: may correspond to current transitions in normal operation, and also depends on the device and gate driver performance, and therefore the range may also be from 10 μs to several hundred microseconds.

[0068] The following steps summarize the action of the protection under abnormal conditions: 1.While Q1 cannot be turned on: a. The control attempts to turn on Q1 (t=t1 under normal conditions). b. The blanking signals toggle from low to high simultaneously (step 1). c. The control sends a signal to Q3 to turn it off at t=t2. d. The blanking signal toggles from high to low and the current derivative signal is not generated during the blanking window. e. A signal in a high state (V BLANK Or V COMP ) causes the protective device to be triggered before a current interruption occurs. 2. While Q1 is turned off and Q3 cannot be turned on: a. The control sends a signal to Q3 to turn it off at t=t1; Q1 continues to conduct DC current. b. The blanking signals toggle from low to high at the same time. c. The control sends a signal to Q1 to turn it off at t=t2. d. After some delay, the current in Q1 slowly decays as Q1 enters avalanche mode starting at t=t3. e. The corresponding current derivative signal is lower than a fixed threshold. f. The blanking signal toggles from high to low. g. A signal in a high state (V BLANK Or V COMP ) is absent, triggering a protective device.

[0069] The disclosed method and device using discrete components that can be placed on a PCB with the switch devices provides extremely fast detection capabilities, especially during turn-on of the reading device. Such devices are particularly useful in terms of switching transition duration.

[0070] The same methods and devices apply to Q2 and Q4 and other transition events between the switching devices in the lower leg of the converter.

[0071] With regard to current differential sensors, the purpose of a dI / dt sensor is to generate a signal proportional to the time rate of change of the collector current of a power switch. When the current IC flowing through the power device is trending from a high to a low value, the output signal is positive. Conversely, when the current IC in the power device is trending from a low to a high value, the output signal is negative. This function can be performed by using a Rogowski coil, which is a wound coil that surrounds the main conductor as shown in Figure 10. Rogowski coils can be integrated within a PCB using several layers of conductors in such a PCB and can be designed to provide a signal of several volts.

[0072] Other types of current differential sensors may also be used.

[0073] The invention is not limited to the examples provided and the same principles can be applied to the switch devices in the lower leg portion as described above, and the output of the protection device can further be connected to the input of the converter controller to provide an alarm signal which can further be sent to a remote controller or management system.

Claims

1. 1. A method of protecting a current-fed converter from overvoltage during an open-circuit fault condition, the current-fed converter comprising a switching cell having two or more legs, each said leg being provided with at least one upper switching device and at least one lower switching device, when an open-circuit fault condition occurs in a switching device in at least one said leg of the switching cell, the method comprising: measuring a current differential signal at the upper switching devices of at least two of the legs or measuring a current differential signal at the lower switching devices of at least two of the legs when a switching transition initiated by a controller of the current-fed converter occurs, the switching transition including a change in conduction state of the upper switching devices and a change in conduction state of the lower switching devices; during at least one of the switching transitions, based on the current derivative signal; when the absolute value of one of the current derivative signals is less than a first predetermined value or is null; or when the sum of the absolute values ​​of the current differential signals is less than a second predetermined value; Triggering protection when either A method comprising:

2. The method of claim 1 , comprising synchronizing measuring the current derivative signal with switching transitions between the switching devices generated by a controller of the current-fed converter.

3. 3. The method of claim 2, wherein the synchronizing includes adjusting a blanking window that starts with a start of the switching transition and ends before, at or after an end of the switching transition, the blanking window inhibiting triggering of the protection during the switching transition.

4. 4. The method of claim 3, comprising determining which device in the switching cell is a leading device and which device in the switching cell is a freewheeling device during the switching transition and measuring the sign of the output voltage across the switching cell to define a width of the blanking window.

5. The method according to any one of claims 1 to 4, wherein triggering the protection comprises providing a current path external to the switching cell to a current source supplying current to the current-fed converter.

6. 6. The method according to claim 1, further comprising, after triggering the protection, recording a fault condition in a controller of the current-fed converter and / or transferring a fault condition of the current-fed converter in a remote controller.

7. An electronic device for carrying out the method according to any one of claims 1 to 6, said electronic device comprising: means for detecting said switching transitions; means for measuring the current derivative signal during the switching transition; an electronic circuit configured to compare the current derivative signal with the first predetermined value or the second predetermined value to trigger the protection; a protection device that, upon triggering of protection by the electronic circuit, ensures an alternative path for a current provided by a current source outside the switching cell; An electronic device comprising:

8. a. the means for measuring the current derivative signal comprises a sensor for sensing a current derivative of a current flowing through each of the legs of the switching cell, and means for conditioning the current derivative signal issued by the sensor; b. the electronic circuit means for providing a first absolute value of the first current derivative signal and means for providing a second absolute value from the second derivative signal; a first comparator that compares the first absolute value with the first predetermined value and provides a first logic output in response to a comparison result; a second comparator that compares the second absolute value with the first predetermined value and provides a second logic output in response to a comparison result; and a logic gate that is connected to outputs of the first comparator and the second comparator and provides a trigger signal that triggers the protection device when either one of the first absolute value and the second absolute value is smaller than the first predetermined value. Or, summing means for providing a sum of the first absolute value and the second absolute value; and a third comparator for comparing the sum with the second predetermined value and providing a third logic output in response to the comparison, and for providing a trigger signal for triggering the protection device when the sum is less than the second predetermined value. Either The electronic device of claim 7 .

9. The electronic device of claim 8 , wherein the sensor for sensing the current derivative is a Rogowski coil.

10. 10. The electronic device according to claim 8 or 9, wherein the means for conditioning the current derivative signal comprises a filter and / or an amplifier.

11. The electronic device according to any one of claims 7 to 10, wherein the means for measuring a current differential signal and the electronic circuit are composed of an analog circuit and a logic circuit.

12. The electronic device according to any one of claims 7 to 11, wherein the protection device comprises a protection switch that freewheels a current supplied to the switching cell.

13. 11. The electronic device of claim 8, further comprising a gate for enabling the trigger signal, the gate having as a first input the trigger signal and as a second input a blanking signal issued from a blanking circuit for suppressing the trigger signal during at least a part of the switching transition, the blanking signal being issued from a controller of the current-fed converter, the gate being configured to suppress the trigger signal during occurrence of the blanking signal.

14. A current-fed converter comprising an electronic device according to any one of claims 7 to 13, said electronic device being provided on a PCB board within said current-fed converter.

Images