Method for monitoring a control device of an electric motor

Abstract

Method for monitoring a control device (150) of an electric motor, wherein a plurality of switching elements (159) are provided in the control device (150), a switching time is measured at at least one of the switching elements (159) and the state of the control device (159) is evaluated on the basis of the measured switching time.

Description

[0001] The invention relates to a method for monitoring a control device of an electric motor and an arrangement for carrying out the method. State of the art

[0002] Various devices for controlling electric motors are known. Such control devices generally comprise an array of switching elements, which is also referred to as an inverter or power amplifier, and comprise at least one half-bridge with a first switching element and a second switching element connected in series with the first switching element. The first switching element is typically referred to as a high-side (HS) switching element, and the second switching element as a low-side (LS) switching element.

[0003] A method for operating a control device for controlling an electric motor is known from the document DE 10 2020 203 016 A1. The control device comprises at least one half-bridge with a first switching element and a second switching element connected in series with the first switching element. The half-bridge is operated with a first dead time during a first switching operation, in which the first switching element is opened and the second switching element is closed. Furthermore, the half-bridge is operated with a second dead time during a second switching operation, in which the second switching element is opened and the first switching element is closed.An output current caused by the operation of the half-bridge and / or an output voltage caused by the operation of the half-bridge is monitored, wherein a first switching characteristic correlated with the turn-off time of the first switching element and / or a turn-on time of the second switching element is determined from the output current and / or the output voltage, and a duration of the first dead time is dynamically adjusted depending on the first switching characteristic. Furthermore, a second switching characteristic correlated with a turn-off time of the second switching element and / or a turn-on time of the first switching element is determined from the output current and / or the output voltage, and a duration of the second dead time is dynamically adjusted depending on the second switching characteristic. In particular, this can also be carried out depending on current, voltage and / or temperature.

[0004] In modern power electronic devices, e.g. inverters or power converter or power control units, such as drivers of electric BDC motors (BDC: brushless DC motor), an electronic output stage is used to drive and control or regulate the currents.

[0005] One of the main problems with power amplifiers is the high error rate during operation, which can be caused by thermal or mechanical stress, electrical overload or aging.

[0006] This always leads to a fault in the power amplifier, with the only possible response to this event being to initiate faulty operation of the power amplifier or to shut down the power amplifier. For safety-critical applications, there is only one possible response after the fault has occurred in the power amplifier, which can potentially endanger the user. Therefore, components are often oversized to minimize the risk of a potential hazard. However, this is associated with high costs.

[0007] Faults in power elements or devices can only be detected if their functionality is impaired. The system can therefore only react to a fault or defect, which can lead to a potential risk for the user. To date, it is not possible to generate a warning for power electronics that the system may soon fail. The only option is to set up a damage counter for the device based on temperature, time, or temperature swings, which can excessively limit its operating time.

[0008] To prevent power amplifier components such as MOSFETs, IGBTs, etc., from overheating and thus being destroyed, a temperature-sensing component, such as a diode on the power amplifier component chip, or a temperature-dependent element such as an NTC, PTC, diode, etc., mounted close to the power amplifier component(s), is used to measure the device's temperature, as most faults result from overtemperature of the chip, which can be detected with such a measure. However, not all faults can be detected by this sensing element, as it passively measures the chip temperature. Therefore, if only a few cells are damaged and cause a cascading fault or failure of the MOSFET, it is unlikely that the sensing element will detect the fault in advance and the system will be able to react appropriately. Disclosure of the invention

[0009] Against this background, a method having the features of claim 1 and an arrangement according to claim 10 are presented. Embodiments emerge from the dependent claims and from the description.

[0010] The presented method is used to monitor a control device of an electric motor, wherein the control device contains a plurality of switching elements, e.g., MOSFETs. A switching time is then measured for at least one of the switching elements. The state of the control device is evaluated based on the measured switching time. This means that the measured switching time allows, for example, a statement to be made about the switching element whose switching time was measured and thus also about the state of the output stage in which the switching element is located, and even about the state of the entire control device. Typically, only one half-bridge can be monitored per signal.

[0011] Based on the fact that temperature and other failure mechanisms lead to increased switching times, the idea is to monitor the switching times of the switching elements. If the switching time measurement versus the measured current and voltage shows a high switching time value, this is an indication that the MOSFET or the MOSFET's thermal cooling path is damaged due to increased temperature. This increased temperature may also be caused by another component.

[0012] Further advantages and embodiments of the invention will become apparent from the description and the accompanying drawings.

[0013] It is understood that the features mentioned above and those to be explained below can be used not only in the combination specified in each case, but also in other combinations or on their own, without departing from the scope of the present invention. Short description of the drawings Fig. 1 shows a graph of signal curves to illustrate the HS turn-off behavior of a damaged MOSFET. Fig. Figure 2 shows a graph of signal curves to illustrate the HS turn-off behavior of an undamaged MOSFET. Fig. 3 shows a circuit diagram of a half-bridge with a fetch path and an entire control device. Fig. Figure 4 shows a circuit diagram of a single HS / LS switch with a retrieval path. Fig. 5 shows a graph of signal curves when measuring the turn-off times of an HS-MOSFET / IGBT. Fig. Figure 6 shows a graph comparing recall signals from a set of recall signals versus temperature. Embodiments of the invention

[0014] The invention is illustrated schematically in the drawings using embodiments and is described in detail below with reference to the drawings.

[0015] Fig. Figure 1 shows the switching behavior of a damaged HS-MOSFET in a graph 10. This is damaged, for example, due to mechanical damage, in this case due to excessive stress and the resulting crack formation in the chip. Graph 10, with time t [s] plotted on the abscissa 12 and voltage [V] on the ordinate 14, shows the U DS 20 on the HS-MOSFET, U DS 22 on the LS-MOSFET, U GS 24 on the HS-MOSFET and U GS 26 shown on the LS-MOSFET.

[0016] Fig. Figure 2 shows the switching behavior of an undamaged HS-MOSFET in a graph 50. In the graph 50, on whose abscissa 52 the time t [s] is plotted and on whose ordinate 54 the voltage [V] is plotted, the curves of U DS 60 on the HS-MOSFET, UDS 62 on the LS-MOSFET, U GS 64 on the HS-MOSFET and U GS 66 shown on the LS-MOSFET.

[0017] As can be seen, the gate plateau voltage of the damaged MOSFET increases, which leads to an increased switching time of the switching device, in this case the MOSFET.

[0018] To measure the switching time, for example, the phase feedback signal or the retrieval path signal can be measured, which can be done with the Fig. 3 shown circuit could be carried out.

[0019] Fig. Figure 3 shows on the left side a half-bridge, which is designated overall by the reference numeral 100. This half-bridge 100 comprises an HS-MOSFET 102 and an LS-MOSFET 104. Furthermore, the illustration shows a connection 110 for the battery voltage V BAT, a terminal 112 for ground, a gate terminal 114 for the HS-MOSFET 102, a gate terminal 116 for the LS-MOSFET 104, a source terminal 118 for the HS-MOSFET 102 and a source terminal 120 for the LS-MOSFET 104. The MOSFETs 102, 104 thus serve as switching elements.

[0020] The diagram also shows an output 122 for the control signal to the load and an output 124 for a state feedback, the so-called call path 124.

[0021] A first curve 130 shows the waveform of a square-wave signal applied to the gate terminal 114 of the HS-MOSFET 104. Another curve 132 shows the waveform of a square-wave signal applied to the fetch path 124. Arrows 140 illustrate a time difference between the two curves 130, 132 during turn-off, and arrows 142 illustrate a time difference between the two curves 130, 132 during turn-on.

[0022] On the right side of the Fig. Figure 3 shows a block diagram of a control device for controlling an electric motor, the control device being designated overall by reference numeral 150. The control device 150 comprises a microcontroller 152, a gate driver unit (GDU) 154, a signal conversion feedback circuit 156, and an output stage 158, in which at least one half-bridge 100 and thus switching elements 159 are provided.

[0023] Signals are a PWM signal U PWM,HS 160 for the HS-MOSFET, a PWM signal U PWM,LS 162 for the LS-MOSFET, a signal U GATE,HS 164 for the gate of the HS-MOSFET, a signal U SOURCE, HS 166 for the source terminal of the HS-MOSFET, a signal U GATE,LS 168 for the gate of the LS-MOSFET, a signal U SOURCE, LS 170 for the source terminal of the LS-MOSFET, a phase signal U PHASE 172, which corresponds to the signal U SOURCE,HS 166 and a signal U PFB, a phase readback signal 174. This is a signal 124 adapted to logic voltage, which can have significantly higher voltages.

[0024] Fig. Figure 4 shows a fetch path for a single high-voltage / low-voltage switch, e.g., for an n-channel MOSFET. A p-channel MOSFET can also be used.

[0025] The diagram shows an LS switch 200 and an HS switch 202 with the connections and control signals as well as signals on the call path as in Fig. 3.

[0026] If a longer switching time is applied to the phase with the circuit, the measuring device, e.g., a microcontroller, can measure the phase's switching time. Thus, if, for example, the THS_OFF time is longer than expected for the provided states, this indicates either an overtemperature or a malfunction of the device.

[0027] Fig. 5 shows different signal curves, namely U GS,HS 250, U GS,LS252, U DS,LS 254, U DS,HS 256, U PWM,HS 258, U PWM,LS 260, U PFB 262. Double arrows indicate time periods, namely a set dead time T pead,set 270, a time T HS,off 272, a time T LS,on 274 and an unused dead time T Dead,unused 276.

[0028] Therefore, the measuring device, e.g., a microcontroller controlling the inverter, is able to detect a fault before it leads to a complete failure of the inverter, and appropriate countermeasures can be taken. The overtemperature can be detected due to the fact that the threshold voltage increases above the temperature, and the resulting increase in the switching time can be measured or estimated, at the same current and voltage, which is shown in the graph in Fig. 6 can be seen.

[0029] Fig.6 shows a comparison between call signals over temperature in a graph 300, on whose abscissa 302 the temperature [°C] is plotted and on whose ordinate 304 the switching time difference [%] is plotted, wherein the course of the determined values ​​is shown together with the course of a linear interpolation.

[0030] The illustration shows a first curve 310 with a difference of 1A and the curve 312 of the interpolation, a second curve 314 with a difference of 2A and the curve 316 of the interpolation, a third curve 318 with a difference of 5A and the curve 320 of the interpolation and a fourth curve 322 with a difference of 10A and the curve 324 of the interpolation.

[0031] The fault or overtemperature can thus be detected if the measured switching time exceeds the calculated switching time at a maximum permissible temperature. For example, a switching time of 850 ns is calculated at a current of 50 A and a drain-source voltage of 13 V at a temperature of 175 °C. If the measured switching time under these conditions is higher than 850 ns, this is an indication that the switching device is experiencing overtemperature or is damaged.

[0032] It is now possible to warn the user of the faulty device, preventing a potentially dangerous situation before the device fails completely. It is also possible to put the device into emergency mode before the defect occurs.

[0033] Some definitions are given below: The switching time is the time period, the propagation delay, between sending a command to the input of the power stage, ie a signal change at the gate, and the time at which the output U DS , ie the drain-source voltage, changes.

[0034] The time between switching the half bridges with: UDS_Highside-off and UDS_Lowside-on is called dead time of the falling edge and UDS_Highside_off and UDS_Highside_on are referred to as the dead time of the rising edge.

[0035] The basic idea is to calibrate the switching time of each MOSFET / IGBT of the power device once at a specified current(s), voltage(s) and chip temperature and to recalculate or determine the switching time(s) of the power device(s) permanently or occasionally during normal operation.

[0036] To evaluate the switching time from the feedback signal, a known current must flow: into the power amplifier to measure the switching on and off time of the circuit breaker, from the power amplifier to measure the on and off time of the HS switch.

[0037] The required known current can be generated by the power amplifier itself or by an external source. The applied voltage can also be measured to increase accuracy. Requirements for the half bridges:

[0038] The dead time between the HV and LS switching is set to more than the required minimum dead time.

[0039] The dead time to the falling edge and the dead time to the rising edge should be determined separately for each half-bridge. Requirements:

[0040] The switching time of the MOSFET / IGBT should be determined at at least one temperature. This can also be done at several temperatures. This measurement(s) is / are used as a reference value.

[0041] The switching times of the MOSFET / IGBT should be determined at least at one current. This can also be done at several currents. This measurement(s) is / are used as a reference value.

[0042] The switching times of the MOSFET / IGBT should be determined at at least one voltage. It is also possible to do this at several voltages. This measurement(s) is / are used as the reference value(s).

[0043] A phase feedback signal should be provided to the logic. This should be done, for example, for a comparator, to make it readable by the internal measuring device, e.g., a microcontroller.

[0044] The differences between the set signal and the read back signal are measured, e.g. with a microcontroller.

[0045] The switching time is the time period, propagation delay, from sending a command to the input of the power amplifier, ie signal change at the gate, until the load voltage changes.

[0046] Advantages of implementations of the procedure compared to previous solutions are: Previous solutions can only detect overtemperature in switching devices by measuring the temperature of the switching element with a separate measuring device.

[0047] Conventional systems can only detect a fault if the device has failed and the system is affected. Now it is possible to detect components that are about to fail.

[0048] The new approach makes it possible to detect errors before they have a disruptive effect on the system.

[0049] This makes it possible to initiate appropriate countermeasures before a malfunction in the system can pose a potential threat to the user.

[0050] Aging effects of the thermal connection between each individual component of the power amplifier and heat sinks can be captured by comparing it with a temperature model over the lifetime.

[0051] Early warnings of the end of operations can be generated.

[0052] Safety margins due to oversizing can be reduced to a minimum value.

[0053] There is no need to implement an error counter that would excessively limit the lifetime.

[0054] Increased security for the user and increased system availability are achieved.

[0055] More failure modes can be detected than when temperature alone is used.

[0056] General advantages of the method regardless of the internal or external sensor used: Digital measurement of feedback signal edges instead of low-level, high-impedance signals from analog thermal sensors is possible, e.g., a parasitic body diode measurement. This is also robust with regard to electromagnetic compatibility.

[0057] The method is applicable to each individual power device in each half-bridge B2, B4, B6, ..., or individual circuit breaker if phase / load feedback monitoring is implemented.

[0058] A simple measuring circuit can be used. QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] DE 10 2020 203 016 A1

[0003]

Claims

[1] Method for monitoring a control device (150) of an electric motor, wherein the control device (150) is provided with a plurality of switching elements (159), a switching time is measured for at least one of the switching elements (159) and the state of the control device (159) is evaluated on the basis of the measured switching time. [2] Method according to claim 1, which is carried out in order to detect a fault in the control device (150). [3] A method according to claim 2, wherein when an error is detected, a user is warned. [4] A method according to claim 2 or 3, wherein, when a fault is detected, emergency operation is initiated. [5] Method according to one of claims 1 to 4, in which a comparison of the measured switching time with another time is carried out. [6] A method according to claim 5, wherein the other time is determined by a calculation. [7] Method according to one of claims 1 to 6, wherein the switching time is determined via a measured polling path signal. [8] Method according to one of claims 1 to 7, wherein the determined switching time is considered over a measured current and a measured voltage. [9] Method according to one of claims 1 to 8, in which MOSFETs (102, 104) are used as switching elements. [10] Arrangement for monitoring a control device (150) of an electric motor, which is designed to carry out a method according to one of claims 1 to 9.

Images