DE102016223511A1 - Method for detecting a failure of a parallel-connected semiconductor - Google Patents

Method for detecting a failure of a parallel-connected semiconductor Download PDF

Info

Publication number
DE102016223511A1
DE102016223511A1 DE102016223511.7A DE102016223511A DE102016223511A1 DE 102016223511 A1 DE102016223511 A1 DE 102016223511A1 DE 102016223511 A DE102016223511 A DE 102016223511A DE 102016223511 A1 DE102016223511 A1 DE 102016223511A1
Authority
DE
Germany
Prior art keywords
signal
parallel
circuit
failure
semiconductor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
DE102016223511.7A
Other languages
German (de)
Inventor
Martin Neuberger
Nils Draese
Mirko Schinzel
Christian Bohne
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Robert Bosch GmbH
Original Assignee
Robert Bosch GmbH
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Robert Bosch GmbH filed Critical Robert Bosch GmbH
Priority to DE102016223511.7A priority Critical patent/DE102016223511A1/en
Publication of DE102016223511A1 publication Critical patent/DE102016223511A1/en
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/26Testing of individual semiconductor devices
    • G01R31/2607Circuits therefor
    • G01R31/2608Circuits therefor for testing bipolar transistors
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/26Testing of individual semiconductor devices
    • G01R31/2607Circuits therefor
    • G01R31/2621Circuits therefor for testing field effect transistors, i.e. FET's
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/26Testing of individual semiconductor devices
    • G01R31/2642Testing semiconductor operation lifetime or reliability, e.g. by accelerated life tests

Abstract

The invention relates to a method for detecting an at least partial failure of at least one semiconductor component which is connected in parallel to other semiconductor components, in which a profile of a supply, which is represented by a measurement signal, a driver, which is provided for driving the semiconductor components, determined and is evaluated and is determined from a post-processing of the course, if at least one of the parallel-connected semiconductor devices has failed.

Description

  • The invention relates to a method for detecting an at least partial failure of at least one parallel-connected semiconductor, in particular in an electrical system of a motor vehicle, and an arrangement for carrying out the method.
  • State of the art
  • Under an electrical system is to be understood in the automotive use, the totality of all electrical components in a motor vehicle. Thus, these include both electrical consumers and supply sources, such as, for example, generators or electrical storage, such as batteries. Furthermore, it also includes all electrical connection and distribution elements such as cable or harness, power distribution and fuse boxes. In the motor vehicle care must be taken to ensure that electrical energy is available in such a way that the motor vehicle can be started at any time and that sufficient power is available during operation. But even when parked, electrical consumers should still be operable for a reasonable period of time, without a subsequent start being impaired.
  • It should be noted that due to the increasing electrification of units and the introduction of new driving functions, the requirement for the reliability of the electrical energy supply in the motor vehicle steadily increases. Furthermore, it must be taken into account that, in the future, highly non-driving activities should be permitted to a limited extent in the case of highly automated driving. A sensory, regulatory, mechanical and energetic fallback by the driver in this case is limited. Therefore, in a highly automatic driving the electrical supply has a previously unknown in motor vehicles safety relevance. Errors in the electrical system must therefore be reliably and completely recognized.
  • Under a highly automatic driving, which is also referred to as highly automated driving, an intermediate step between an assisted driving in which the driver is assisted by assistance systems, and an autonomous driving in which the vehicle drives automatically and without the driver's intervention, to understand. In the case of highly automatic driving, the vehicle has its own intelligence that could plan ahead and take on the driving task, at least in most driving situations. Therefore, in a highly automatic driving the electrical supply has a high safety relevance.
  • Multiphase DC-DC converters have long been known and are also used in multiple-voltage on-board networks. The parallel connection of several semiconductors is already known as well as the control via a driver module. The diagnosis of the entire DC-DC converter is known. The testing of primary windings of voltage transformers is also known. In addition, there are methods in which the switching converter can be diagnosed via FFT (Fast Fourier Transformation) or temperature measurements.
  • Also known are drivers for MOSFETs, which have already implemented a detection via the switching behavior.
  • Disclosure of the invention
  • Against this background, a method with the features of claim 1 and a circuit arrangement according to claim 11 are presented. Embodiments emerge from the dependent claims and from the description.
  • The presented method is intended to detect the failure of a semiconductor component or of a semiconductor component which is connected in parallel with other components or components, for example of the same type. In particular, a semiconductor device of the kind discussed herein is a power semiconductor device, such as a MOSFET or an IGBT, connected in parallel in an electronic device, such as a DC-DC converter. In the method, it is now provided to measure a supply current and thereby to determine the level of the supply current or its charge quantity, which in turn is proportional to a number of semiconductor semiconductors or power semiconductors to be triggered.
  • In this way, a single failure can be detected in an embodiment of an array of MOSFETs connected in parallel. This means that even if one of four parallel-connected semiconductor devices fails or is interrupted, and thus the three remaining ones can still maintain the function of the circuit, this failure can be detected. This is of particular importance, since in later operation the failure of the remaining semiconductor components, such as semiconductor switches, threatens. The method thus enables error detection for various components, such as MOSFETs or IGBTs, in a parallel arrangement.
  • In an embodiment, it is also possible to detect the exact number of failed semiconductor, regardless of the basic number of parallel interconnected semiconductors. Thus, the method is, for example. applicable in a parallel connection of four MOSFETs.
  • Furthermore, it is possible in the embodiment to detect the failure of a single phase of a DC-DC converter during operation in multi-phase circuits. In addition, it is possible to determine the failed phase and a balancing of the remaining phases to prevent overheating and thus shutdown. This can be used, for example, in the context of implementing an emergency operation.
  • The presented method enables the detection of the partial failure of clocked parallel-connected MOSFETs over the time course of the supply current of the driver. Since usually the parallel-connected MOSFETs are driven by a common driver and parasitic capacitances must be loaded or reloaded at each driving process, the supply current or its peak of the driver behaves in proportion to the number of semiconductor devices to be driven. This peak supply current serves as the basis of the diagnosis.
  • In this connection, a method for detecting the aforementioned failure and a circuit arrangement for carrying out the method are presented in particular.
  • The presented circuit arrangement for detecting a failure can be used in an embodiment regardless of the amount of voltage supply of the driver or drive module. Likewise, the required positive supply voltage of the signal conditioning circuit can be provided by an additional circuit.
  • An advantage of the presented method, at least in some of the embodiments, is that in multi-phase systems with sequential control only one evaluation circuit is necessary for all phases. Upon detection of the partial failure of parallel semiconductor switches in multi-phase converters, a limp home mode can be realized by the current load of the partially defective phase is minimized by a phase current control.
  • It is important that even a complete phase failure can be detected. By adaptations in the control, a run-flat operation can be realized by the remaining phases are controlled so that an overload of the remaining phases is prevented.
  • Even with single-phase systems, complete failure can be preventively prevented by interruption if it is detected that individual semiconductors have failed in a parallel circuit.
  • The developed circuit arrangement makes it possible to diagnose the phase failure. For this purpose, for example, a measuring resistor is implemented in the central power supply for all phase drivers, ie a signal conditioning circuit for all phases. Since in the embodiment is a nested multiphase converter with a time-offset timing, which is the rule, caused by the driver voltage peaks or peaks of the individual phases are offset in time. For this reason, only a single implementation of the measurement circuit, ie only one circuit arrangement, is needed to diagnose all phases. In the case of multiphase converters, the phases are always switched in an offset manner, otherwise some advantages of the topology would be omitted. The signal curve is visible on the measuring resistor used. It will open 9 and 10 directed.
  • The presented method also diagnoses the phase failure caused by the failure of all phase MOSFETs of multiphase transducers. For this purpose, the threshold value is set to a low value, which is used to diagnose a failure of all MOSFETs, for example 10% to 15% of the maximum peak ripple of the phases.
  • The circuit arrangement and the method can be used for different evaluations. In an interrupt evaluation, for example, the threshold value of the comparator (see 6 ) is set so that at partial failure a continuous high or continuous low level is present at the microcontroller. This method does not require high computing power, as it only responds to toggling of the output (see 7 and 8th ).
  • For an analog average or average evaluation, a comparator output (see 6 ) is connected to an ADC pin of the microcontroller and the average voltage value is evaluated. If the average voltage drops below a threshold defined in the microcontroller, a certain number of MOSFETs have failed. Thus, the exact number of defective components can be diagnosed.
  • In a digital scan, the comparator output (see 6 ) connected to a digital signal processor. The signal is sampled at a multiple of the MOSFET switching frequency and a high / low evaluation is performed. Due to the 0-1 evaluation, small influences occur. Digital signal processors also provide fast processing of the signal.
  • Further advantages and embodiments of the invention will become apparent from the description and the accompanying drawings.
  • It is understood that the features mentioned above and those yet to be explained below can be used not only in the particular combination indicated, but also in other combinations or in isolation, without departing from the scope of the present invention.
  • list of figures
    • 1 shows a block diagram of an embodiment of a dual-channel vehicle electrical system according to the prior art.
    • 2 shows a simplified representation of a four-phase up-down switching converter according to the prior art.
    • 3 shows a failure of a parallel MOSFET.
    • 4 shows measuring points of a driver monitoring.
    • 5 shows a signal conditioning circuit.
    • 6 shows a PWM threshold and a comparator circuit.
    • 7 shows signal curves in error-free case ..
    • 8th shows waveforms in case of error.
    • 9 shows voltage ripple on the measuring resistor.
    • 10 shows an output of a comparator in an interleaved multiphase DC-DC converter.
  • Embodiments of the invention
  • The invention is schematically illustrated by means of embodiments in the drawings and will be described in detail below with reference to the drawings.
  • The invention is illustrated by the example of a 48 V / 14 V voltage converter, but is also applicable to many other electronic components, even in pure 14 V, 48 V or high-voltage on-board networks.
  • 1 shows a wiring system, the whole with the reference number 10 is provided and represents a dual-channel electrical system according to the prior art. There is a base board network 11 marked with a border in which 48V components and 14V components without safety relevance are provided. Furthermore, a first channel 13 and a second channel 15 intended.
  • The illustration shows a generator or an electrical machine 12 , which supplies a voltage of 48 V, a first so-called electronic power distribution unit 14 (electronic energy supply unit, ePDU), a first consumer 16 , a first battery 18 with 48V, a first battery management system 20 (BMS: Battery Management System), a first DC-DC converter 22 , which converts the voltage of 48 V into a voltage of 14 V, a second ePDU 24, a second consumer 26 , a second battery 28 with 14V with a battery sensor 30 , a second DC-DC converter 32 , which converts the voltage of 48 V into a voltage of 14 V, a third battery 34 with a battery sensor 36 , a safety-relevant consumer Rs1a 38, the function of which is redundantly fulfilled by a consumer Rs1b 40, a safety-relevant consumer Rs2a 42 having an internal, redundant consumer Rs2b 44.
  • In the base board network 11 are therefore two batteries 18 and 28 intended. The first safety-relevant channel 13 is to the base board network 11 coupled and includes safety-relevant consumers, such as brake and steering. The second safety-relevant channel 15 also contains safety-relevant consumers. Since also in this safety-relevant component are supplied with 14V, are the second DC-DC converter 32 and the third battery 34 intended.
  • 2 shows in a block diagram a simplified representation of a four-phase DC-DC converter, in total with the reference numeral 100 is designated and designed for both an up and down converter operation. The illustration shows a high-side 102 with four parallel-connected MOSFETs 104 and a low-side 106 also each with four MOSFETs connected in parallel 108 , Furthermore, a ferrite module 120 , a circuit breaker 122 for the high side 102 , a circuit breaker 124 for the low side 106 , a first filter 130, a second filter 132 , Clamp 40 134 , Clamp 30 136 as well as clamp 31 138 reproduced.
  • So far, today only the complete failure of the DC-DC converter 100 be recognized. Even currently developed methods are unable to determine which phase has failed. Therefore, it is not possible to detect the failure of individual semiconductors without a special driver in parallel circuits of semiconductor devices. Such a parallel connection is in 3 shown.
  • 3 shows a parallel connection on the left side 150 of four MOSFETs 152, 154, 156, 158 as semiconductor elements or semiconductor switches. Furthermore, stream arrows are entered, namely for a inflowing current I to 160, which divides an outgoing current I from 162 and into four partial flows I 1 170, I 2 172, I 3 174 and I 4 176.
  • On the right side is also a parallel connection 250 of four MOSFETs 252, 254, 256, 258. Furthermore, current arrows are entered, namely for an incoming current I to 260, the one outgoing current I from 262 and into four partial currents I 1 270, I 2 272, I 3 274 and I 4 276 splits. Because the left MOSFET 252 has failed, the current I 1 = 0.
  • Since, according to the Kirchhoff node rule, the sum of the incoming currents is equal to the sum of the outgoing currents, the remaining MOSFETs must 254 . 256 . 258 the current of the failed MOSFET 252 take. This means that the inflowing current I to 260 is divided into the three partial currents I 2 272, I 3 274 and I 4 276. This in turn has a temperature increase of the remaining MOSFETs 254 . 256 . 258 This results in an increase of the internal resistance RDS (ON) due to the positive temperature coefficient of MOSFETs. In addition, this increases the losses and the temperature of the circuit. Thus, a subsequent failure becomes more likely.
  • The object of the invention is the detection of a single failure of an array of parallel MOSFETs. This means that even if one of four parallel-connected semiconductor devices one failed or interrupted and thus the remaining three can still maintain the function of the circuit, this failure is to be detected, since in later operation the failure of the remaining semiconductor switch threatens.
  • 4 shows a driver 300 for parallel-connected semiconductor components, in particular of semiconductor switches, such as, for example, MOSFETs or IGBTs. The illustration shows a supply voltage driver 302 , a measuring resistor 304 , a blocking capacitor 306 , an inductance 308 , a zener diode 310, a capacitance 312 , a first pulse driver 314 and a second pulse driver 316 , The two pulse drivers 314 . 316 each set a PWM signal (PWM: pulse width modulation) to control the driver 300 via a first entrance 320 and a second entrance 322 ready. About a third entrance 324 the supply, in this case a supply voltage, is provided. This supply and in particular its course can now over the current, by the measuring resistance 304 flows, or via the voltage at the blocking capacitor 306 be determined. For this purpose, a first measuring point 330 and / or a second measuring point 332 be provided.
  • The supply and in particular its course can thus either via a course of the supply current at the measuring resistor 304 , which is connected in series to a supply line, or with the blocking capacitor 306 or a backup capacitor on the driver 300 be determined. This backup capacitor is typically located between the supply pins of the driver 300 and the ground pin of the driver 300 ,
  • The driver is at an output 340 Signals for gate connections of high-side MOSFETs, at a second output 342 Signals for source connections of high-side MOSFETs and drain connections of low-side MOSFETs and at a third output 344 Signals for gate connections of low-side MOSFETs.
  • The invention is explained using the example of a voltage converter for a boost recuperation system, but applies to all circuits with parallel-connected semiconductor components or power semiconductors which are operated in a clocked manner. The voltage converter is a 4-phase DC switching converter, which can convert 14 V to 48 V or 48 V to 14 V. The diagnosis basically works with control units or components which contain MOSFET or IGBT drivers with external semiconductor switches, such as, for example, MOSFETs or IGBTs.
  • The voltage converter is four-phase, each phase has four parallel high-side and four parallel-connected low-side MOSFETs. Each phase also has a respective driver component, which alternately controls the low-side and high-side MOSFETs.
  • One way of measuring the current consumption of the MOSFET drive module is the measurement of the voltage ripple via a defined resistor which is connected in series with the supply line, in 4 this is the first measurement point 330 , This defined resistance is referred to below as a measuring resistor or shunt. In the proposed method, a standard SMD resistor with a tolerance range of + -1% and a resistance of Rshunt = 1 Ω can be used. The current required by the driver chip flows through the resistor causing a voltage drop across the resistor. About the voltage drop, an indirect current measurement is performed and evaluated in the next step in a developed measurement circuit.
  • In an alternative embodiment, the current of the driver supply is determined by measuring the voltage drop across a blocking capacitor.
  • In a further embodiment, the low-cost SMD resistor (shunt) can be replaced by a high-precision measuring shunt for increased accuracy.
  • The use of operational amplifiers (OPV) for measurement signal conditioning offers several advantages. Which includes:
    • high input resistance of OPVs in the range of MΩ to GQ, resulting in a negligible influence on the circuit to be measured,
    • - Small bias currents, these are in bipolar OPVs at 10 nA to a few 100 nA, in FET input stages, these values are close to 0 A,
    • - Reaction time is extremely fast due to the analog circuit technology used.
  • 5 shows a signal conditioning circuit 400 used to remove a DC offset and to amplify the signal.
  • A subtractor 402 consisting of resistors 404 . 406 , an operational amplifier 408 , Resistors 410 . 412 Towards ground and a resistor 414 parallel to the operational amplifier 408 the signal conditioning circuit 400 is used to remove the DC offset, this is the signal conditioning circuit 400 can be used regardless of the level of the driver supply voltage. By removing the DC offset as possible voltage fluctuations in the supply line of the drive module for the diagnosis are not critical.
  • A successive amplifier link 450 that is an operational amplifier 452 and resistances 454 and 456 includes, serves to prepare the basic signal. The signal conditioning circuit 400 does not require a separate positive supply to the operational amplifier 408 and 452 because they provide themselves through the circuit structure. An advantage of analog signal conditioning is the high speed of error detection. Furthermore, the component costs are low.
  • The subsequent signal processing of the output signal of the amplifier or the amplifier element 450 alternatively, it can be done by means of a microcontroller, digital signal processor (DSP) or field programmable gate array (FPGA) (digital or analog). In this case, the voltage drop is queried via an ADC input PIN and evaluated integrally via a change in the voltage surface or evaluated by means of an internal comparator.
  • 6 shows an example of signal conditioning from a PWM output value 480 above a threshold 482 , This threshold 482 can z. B. via a microcontroller and the PWM. This results in an output signal, the input signal 502 for a comparator 490 together with the output signal 500 the in 5 illustrated circuit arrangement. The values 500 and 502 are compared with each other. Thus, it is by means of the comparator or the comparator circuit 490 possible, the prepared measuring signal 500 or to evaluate the diagnostic signal. By indirect current measurement via a measuring resistor or shunt 504 and the freely definable threshold 502 the comparator evaluation takes place. The threshold value can be set permanently via a high-impedance voltage divider or alternatively via a PWM output of a microcontroller. When setting the threshold value via a PWM signal, an initial run of the diagnostic concept is also possible. Depending on the type of connection of the OPV, the output of the comparator switches between its positive or negative modulation range when it exceeds or falls below the defined threshold value. The evaluation of the output signal of the comparator can be done in various ways.
  • A possible output signal of the analog signal conditioning in connection with the comparator circuit is in 7 and 8th shown.
  • 7 shows the waveform of the voltage at the output of the signal conditioning circuit in the error-free case, ie all MOSFETs are operational. The illustration shows in a graph 600 , on the abscissa 602 the time [μs] and at its ordinate 604 the voltage [V] is plotted, the voltage curve 610 , the course of a signal 612 to the microcontroller and a level 614 a threshold that is individually adjustable.
  • The pulsed voltage curve 610 with a significant peak value is attributed to the current flow through the recharge of the gate charges in the MOSFET or IGBT: To open or close a semiconductor switch such as a MOSFET, the parasitic gate capacitances in the silicon must be charged or reloaded. A power is needed for this reloading. This initially increases sharply and then flattens off more and more. Since a large current flows when charging a capacitor, it then levels off with increasing charging status. This behavior reflects the general charge / discharge curve of a capacitor
  • As long as the signal 610 the threshold 614 exceeds, is the digital signal 612 = 1, otherwise 0. The threshold value can therefore be set so that only when all parallel-connected gate charges are charged, the threshold is exceeded,
    8th shows the waveform of the voltage at the output of the comparator in the faulty case, ie one or more parallel power semiconductors are, for example. Defective. The illustration shows in a graph 700 , on the abscissa 602 the time [μs] and at its ordinate 604 the voltage [V] is plotted, the voltage curve 710 , the course of a signal 712 to the microcontroller, in the faulty case a permanent low signal on the microcontroller, and a level 714 a threshold that is not exceeded in this case.
  • In error-free operation according to 7 Depending on the set switching frequency of the MOSFETs an equally clocking high level 612 output from the comparator. At a partial failure of 50% of the parallel MOSFETs, as in 8th is shown, the threshold value 714 not reached and a continuous LOW signal 712 is at the microcontroller.
  • The evaluation of the comparator signal (reference numeral 612 / 712 in 7 respectively. 8th ) can be realized in different ways. By interchanging the input pins of the comparator, the output signal can always be inverted. The positive as well as negative level of the output voltage depends on the supply voltage of the OPV. The evaluation of the diagnosis or the output signal (reference numeral 612 / 712 in 7 respectively. 8th ) can be realized with the following approaches:
    • - Interrupt evaluation: Here, the threshold value of the comparator is set so that at partial failure a continuous high or continuous low level is present at the microcontroller.
  • Advantage: This method does not require a high computing power, as it only responds to toggling of the output (see 7 and 8th ).
    • - Analog average evaluation: The comparator output 612 / 712 is connected to an ADC PIN of the microcontroller and the average voltage value is evaluated. If the average voltage drops below a threshold defined in the microcontroller, a certain number of MOSFETs have failed.
  • Advantage: The exact number of defective components can be diagnosed.
    • - Digital Sampling: The comparator output 612 / 712 is connected to a digital signal processor (DSP), the signal is sampled at a multiple of the MOSFET switching frequency and a high / low evaluation is performed.
  • Advantage: Due to the 0-1-evaluation low influences, DSPs offer a fast further processing of the signal.
  • Especially for multi-phase components, such as the 48V-14V voltage transformer ( 2 ), the invention can be further developed as follows.
  • 9 shows the waveform in which a four-phase transducer is diagnosed using the above-mentioned facilities. 9 shows in a graph 800, on the abscissa 802 the time and its ordinate 804 the tension is applied, the course 810 the voltage drop at the shunt for the first phase 820 , the second phase 822 , the third phase 824 and the fourth phase 826 for the error-free case. An additional registered history 830 in the second phase 822 illustrates an error 832 in this phase. The history 830 remains below the variably adjustable PWM threshold 840 , 9 thus shows the input signal, corresponding to 610/710 in the previous figures for the iO case and the niO case in one picture.
  • 10 Fig. 12 also shows the output signal of the comparator circuit corresponding to 612/712 in the previous figures for the case of iO and the case of ni0 in an image. 10 shows in a graph 900 , at the abscissa 902 the time and at its ordinate 904 the tension is applied, the course 910 the pending signal on the microcontroller when the diagnosis is applied to a multiphase converter for the first phase 920 , the second phase 922, the third phase 924 and the fourth phase for the error-free case. Another course 930 shows this signal in case of an error 932 in the second phase.
  • The developed circuit arrangement makes it possible to diagnose the phase failure. For this purpose, for example, the measuring resistor is implemented in the central power supply for all phase drivers, ie a signal conditioning circuit for all phases. Since it is an interleaved multiphase converter with staggered timing of the phases, which is the rule, the current peaks caused by the driver of the individual phases are offset in time. For this reason, only a single implementation of the measurement circuit is needed to diagnose all phases. It should be noted that in multi-phase converters, the phases are shifted in almost all designs, otherwise some advantages of the topology would be omitted. The waveform is off at the measuring resistor used 9 seen. In error-free operation ( 9 and 10 , Reference numbers 810 . 910 ) results in the signal waveform shown.
  • The threshold shown 840 for the comparator can be realized by means of a PWM output of the microcontroller, thereby the value can be variably adjusted. With reference numbers 830 respectively. 930 is the waveform of the fault, in this case a partial failure of MOSFETs in phase 2 represented.
  • Due to the decrease in the peak current of the control module of phase 2 below the defined threshold 840 the level of the comparator changes to low in phase 2 ,
  • The presented method also diagnoses the phase failure caused by the failure of all phase MOSFETs by multiphase transducers. For this purpose, the threshold value is set to a low value, which is used to diagnose a failure of all MOSFETs, for example 10% -15% of the maximum peak ripple of the phases.
  • Depending on the level of the variable comparator threshold of the microcontroller, the partial failure of MOSFETs or the failure of a whole phase can be detected. To detect a phase failure, the threshold value would have to be set very low, for example in the range of 10% of the peak voltage value of the phases. This is due to the fact that in case of failure of a phase of the drive module does not receive power. The failure of a phase can have various causes, by means of this diagnosis, the phase failure due to the failure of the driver chip of the parallel-connected MOSFETs or interruption of the driver PWM signal from μC can be detected. In this method too, during an initialization run of the converter, the maximum voltage peak value can be determined in order to take into account possible influences of the temperature or component tolerances for the diagnosis.
  • Since the individual phases of the converter are controlled offset in time by the PWM signal of the microcontroller and the microcontroller is also the information of a partial failure / complete failure of a phase by the returned comparator signal of the measuring circuit is available, by an internal software comparison between the outgoing PWM signal for controlling the driver and the diagnostic signal of the measuring circuit an exact determination of the failed phase possible. By informing the exactly failed phase number, a symmetrization of the functional phases can be carried out. As a result, an emergency operation of the converter can be provided. An example of this would be an increase in the duty cycle of the phases 1 . 3 and 4 Partial failure of MOSFETs in phase 2 , This reduces the current load from phase 2 and consequently the life of the remaining MOSFETs is increased.
  • The threshold 840 of the comparator (reference numeral 490 in 6 ) is incrementally increased upon booting of the component until the peak value of the conditioned measurement signal is detected. If the threshold value exceeds the processed measuring signal, the comparator output tilts and the peak value of the measuring signal is detected. On the basis of this maximum value becomes individually the threshold value 840 set. With this method, it is possible to minimize external influences on the diagnosis, since thus influences, such as the ambient temperature or EMC radiation, can already be fundamentally taken into account in the initialization run. The problem of a faulty initialization run due to an already existing unknown partial failure of MOSFETs before or during initialization can be neglected. This can be justified by the fact that the diagnosis is continuously active as soon as the converter is in operation.
  • The method can be used for monitoring all converters with parallel-connected semiconductor or multi-phase converters. In particular, in the transducers in which an increased self-diagnosis and emergency running properties are required, eg. B. in an automated driving or limp home operation.
  • The method may also provide that an initialization run is performed during a boot process that sets the thresholds of the comparator circuit. As a result, external environmental influences, such as. Temperature, EMC radiation, component tolerances, etc., minimized, so that the diagnostic accuracy and thus their quality is increased.
  • The method can also be used with other products than those described herein, in which high currents are switched and thus parallel semiconductors are installed.

Claims (13)

  1. Method for detecting an at least partial failure of at least one semiconductor component which is connected in parallel with other semiconductor components, in which a profile of a supply, which is represented by a measurement signal, of a driver (300), which is provided for driving the semiconductor components, determined and evaluated is determined and from a post-processing of the course, whether at least one of the parallel-connected semiconductor devices has failed.
  2. Method according to Claim 1 in which at least one semiconductor component is a semiconductor switch.
  3. Method according to Claim 1 or 2 in which the course of the supply current is measured with a measuring resistor (304, 504) nested in series with a supply string.
  4. Method according to Claim 1 or 2 in which the course of the supply is measured with a blocking capacitor (306) on the driver.
  5. Method according to one of Claims 1 to 4 in which a measurement signal, which indirectly represents the measured profile of the supply current, is processed by an operational amplifier before an evaluation.
  6. Method according to one of Claims 1 to 5 in which the measuring signal is conditioned by a signal conditioning circuit (400) before an evaluation, with which a DC offset is removed and an amplification of the measuring signal is carried out and subsequently the mean value, the root mean square or the maximum height of the signal is determined.
  7. Method according to one of Claims 1 to 6 in which, for evaluation, the measurement signal is compared with a threshold value (482), in particular by means of a threshold value formed by a PWM signal, by means of a comparator circuit (490) which outputs a signal which is used for the evaluation.
  8. Method according to one of Claims 1 to 7 in which, during a boot process, an initialization run is performed which determines the thresholds (482) of the comparator circuit (490).
  9. Method according to one of Claims 1 to 8th in which, in the case of a circuit with parallel, successively connected phases, a failure of a parallel-connected semiconductor component with only one circuit arrangement is detected.
  10. Method according to one of Claims 1 to 9 in which, in the case of a circuit with parallel, successively connected phases, the failure of a complete phase is detected.
  11. Circuit arrangement for detecting an at least partial failure of at least one semiconductor component which is suitable for carrying out a method according to one of the Claims 1 to 10 is set up.
  12. Circuit arrangement according to Claim 11 having a comparator circuit (490) whose output signal is to be used for evaluation.
  13. Circuit arrangement according to Claim 11 or 12 which is associated with a signal conditioning circuit (400).
DE102016223511.7A 2016-11-28 2016-11-28 Method for detecting a failure of a parallel-connected semiconductor Pending DE102016223511A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
DE102016223511.7A DE102016223511A1 (en) 2016-11-28 2016-11-28 Method for detecting a failure of a parallel-connected semiconductor

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102016223511.7A DE102016223511A1 (en) 2016-11-28 2016-11-28 Method for detecting a failure of a parallel-connected semiconductor
PCT/EP2017/079708 WO2018095835A1 (en) 2016-11-28 2017-11-20 Method for detecting a failure of a semiconductor connected in parallel

Publications (1)

Publication Number Publication Date
DE102016223511A1 true DE102016223511A1 (en) 2018-05-30

Family

ID=60452635

Family Applications (1)

Application Number Title Priority Date Filing Date
DE102016223511.7A Pending DE102016223511A1 (en) 2016-11-28 2016-11-28 Method for detecting a failure of a parallel-connected semiconductor

Country Status (2)

Country Link
DE (1) DE102016223511A1 (en)
WO (1) WO2018095835A1 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111679170A (en) * 2020-06-09 2020-09-18 浙江大学 Transistor array structure design method based on reliability rapid test

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102007024175A1 (en) * 2007-05-24 2008-12-11 Robert Seuffer Gmbh & Co. Kg Early failure detection method for e.g. metal oxide semiconductor FET, involves measuring direct current portion or direct current resistance at control electrode, and evaluating portion to determine failure tendency of arrangement
JP5983274B2 (en) * 2012-10-09 2016-08-31 富士電機株式会社 Gate drive circuit having failure detection circuit of semiconductor switch element

Also Published As

Publication number Publication date
WO2018095835A1 (en) 2018-05-31

Similar Documents

Publication Publication Date Title
JP4677038B2 (en) Ground fault detection system and ground fault detection method
US9581634B2 (en) Power supply current monitoring device
US8334670B2 (en) Method and apparatus to monitor an electric motor control circuit
US8929113B2 (en) Capacitor discharger for power conversion system
US9944188B2 (en) On-board electrical system and method for operating an on-board electrical system
KR102155973B1 (en) Half-bridge driver fault diagnosis system and method
DE102014114716A1 (en) Monitoring device for a power supply current
US20130039107A1 (en) Discharge circuit for capacitor
JP5808678B2 (en) Electric vehicle
JP6697869B2 (en) State determination device and state determination method
US10530250B2 (en) Multiphase converter
US9140760B2 (en) Resistance measurement for battery cells
JP4638914B2 (en) Electrical circuit for drive control of piezoelectric elements, in particular piezoelectric elements of automobile fuel injection devices
KR101949509B1 (en) An Apparatus and A Method For Testing A Motor Driving IC
DE102016223511A1 (en) Method for detecting a failure of a parallel-connected semiconductor
CN105785136B (en) Direct current link capacitance measurement for electric vehicle drive train
US20200023741A1 (en) Dc-to-dc voltage converter, voltage supply device, and diagnostic method for a dc-to-dc voltage converter
CN106464196B (en) With the control equipment powered off safely
WO2020029901A1 (en) Method for detecting an isolation fault
US9182442B2 (en) Semiconductor module
US10622685B2 (en) Problem detection apparatus
Ashourloo et al. Robust fault detection of hybrid switched-capacitor dickson converter for fault-tolerant automotive applications
US9350229B2 (en) System and method of controlling an inverter based on a calculated capacitance of a DC link capacitor
JP2017055620A (en) Power converter
JP2004515999A (en) Device for identifying errors in a multi-voltage power supply network