Method for detecting wire bond peel of IGBT module based on surface magnetic signal

Abstract

The present application belongs to the technical field of IGBT bonding wire fault detection, and specifically provides an IGBT module bonding wire peeling detection method based on surface magnetic signals, to avoid the invasiveness and destructiveness of the existing bonding wire peeling detection method based on magnetic signals due to sensor arrangement. The present application uses a ferrite coil magnetic probe to capture a significant voltage signal induced by the transient magnetic field on the surface of the IGBT module excited by the load current in the off process, analyzes the response characteristics of the bonding wire peeling event based on the induced voltage signal, and finally realizes IGBT module bonding wire peeling detection, completes non-invasive detection without electrical direct connection, and has wide application prospects in in-situ online detection of bonding wire peeling.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of IGBT bonding wire fault detection, and specifically provides an IGBT module bonding wire peeling detection method based on a surface magnetic signal. BACKGROUND

[0002] Insulated Gate Bipolar Transistor (IGBT) packaging modules are widely used in smart grids, electric vehicles, and high-speed railway transportation systems as key switching devices due to their large capacity, fast switching speed, easy driving, low on-state voltage drop, and high input impedance. As a key component of IGBT module packaging, bonding wires play a crucial role in electrically connecting IGBT chips with external pins of the module. However, due to the mismatch in thermal expansion coefficients between the bonding wires and the silicon chips, and the temperature fluctuations caused by power cycling during the operation of the IGBT module, the bonding wires are prone to peeling off from the chips, which seriously threatens the reliability of the IGBT packaging module. Therefore, implementing bonding wire peeling detection is of great significance for fault warning and prevention of IGBT packaging modules.

[0003] Currently, existing bonding wire peeling detection methods are mainly based on three types of physical parameters sensitive to bonding wire peeling: junction temperature, electrical parameters, and magnetic signals. The first type is based on junction temperature detection, which is achieved through optical non-contact measurement or thermal parameter detection, but is affected by multiple factors such as environmental temperature and working conditions, making this detection method less specific to bonding wire peeling. The second type is based on detection of various electrical parameters, such as bonding wire aging voltage, self-excited short-circuit current, and stray inductance, but this method usually requires a direct electrical connection between the detection system and the module. Considering the strong load current and high voltage working conditions of the IGBT module, the detection system inevitably faces the risk of failure caused by the surge and overshoot voltage of the IGBT module. The third type is based on magnetic signal detection, which uses the change of the induced magnetic field near the current-carrying bonding wire as a fault indication, effectively avoiding the electrical connection problem, and making this method more specific to bonding wire peeling due to the temperature-insensitive nature of magnetic signals. Unfortunately, current bonding wire peeling detection methods based on magnetic signals still face unresolved bottlenecks. Specifically, to capture weak magnetic signals, the sensor needs to be placed very close to the bonding wire, usually requiring the disassembly of the IGBT module packaging structure, including removing the plastic shell and cleaning the silicone filler, which is obviously invasive and destructive. SUMMARY

[0004] The application aims to provide an IGBT module bonding wire peeling detection method based on surface magnetic signals, so as to avoid the invasiveness and destructiveness caused by sensor arrangement in the prior art.

[0005] To achieve the above object, the technical scheme adopted by the application is as follows:

[0006] An IGBT module bonding wire peeling detection method based on surface magnetic signals comprises the following steps:

[0007] S1, a finite element method is used to model and simulate the IGBT module to be detected, the side surface closest to the bonding wire to be detected in the geometric model of the IGBT module to be detected is marked as a detection surface, a single bonding wire peeling is simulated in the simulation process, the rate of change of the normal magnetic flux density of the detection surface under the condition that each bonding wire is peeled alone is obtained, a rectangular sensitive detection area is demarcated according to the rate of change of the normal magnetic flux density of the detection surface, and the rectangular sensitive detection area covers the area where the rate of change of the normal magnetic flux density is greater than a preset sensitive threshold under the condition that each bonding wire is peeled alone;

[0008] S2, a double-pulse test circuit is built for the IGBT module to be detected, and a ferrite coil magnetic probe is used to scan the rectangular sensitive detection area on the surface of the IGBT module to be detected in the test process, so as to obtain the initial induced electromotive force distribution of the rectangular sensitive detection area;

[0009] S3, the same test process in S2 is maintained, each bonding wire of the IGBT module to be detected is cut, the induced electromotive force distribution of the rectangular sensitive detection area under the condition that a single bonding wire is peeled is measured, the induced electromotive force distribution is subtracted from the initial induced electromotive force distribution, and the induced electromotive force rate of change distribution of the rectangular sensitive detection area under the condition that each bonding wire is peeled alone is calculated;

[0010] S4, the induced electromotive force rate of change distribution of the rectangular sensitive detection area under the condition that each bonding wire is peeled alone is fused, for each scanning point, the minimum value of the induced electromotive force rate of change is taken as a bonding wire peeling representation value, so as to form a bonding wire peeling representation distribution;

[0011] S5, sort the characteristic values in the bonding wire peeling characterization distribution according to the order from large to small, select the scanning points corresponding to the first N characteristic values as the detection points, and select the minimum value of the bonding wire peeling characterization value in all the detection points as the alarm threshold;

[0012] S6, fix the ferrite coil magnetic probe on the detection points of the IGBT module to be detected, and capture the induced electromotive force in the working process of the IGBT module to be detected in real time. If the rate of change of the induced electromotive force of any ferrite coil magnetic probe exceeds the alarm threshold, it is determined that the bonding wire peeling occurs.

[0013] Further, the specific process of step S1 is:

[0014] S11, for the IGBT module to be detected, an IGBT geometric model is established in SOLIDWORKS;

[0015] S12, the IGBT geometric model is imported into a finite element analysis software, the side surface of the IGBT geometric model closest to the bonding wire to be detected is marked as a detection surface; the IGBT module to be detected is first subjected to finite element meshing to obtain a meshing graph; then the IGBT geometric model is subjected to electric field and magnetic field, and an excitation condition of static load current is added in the electric field to complete electromagnetic simulation, and an initial normal magnetic flux density distribution of the detection surface is obtained;

[0016] S13, under the same simulation conditions in S12, the conductivity of each bonding wire is set to 10 -12 S / m to simulate the peeling of a single bonding wire, and the normal magnetic flux density distribution of the detection surface under the peeling state of each bonding wire is obtained, and the rate of change of the normal magnetic flux density of the detection surface under the peeling state of each bonding wire is calculated;

[0017] S14, a rectangular sensitive detection area is demarcated according to the rate of change of the normal magnetic flux density of the detection surface, and the rectangular sensitive detection area covers the area where the rate of change of the normal magnetic flux density under the peeling state of each bonding wire is greater than a preset sensitive threshold.

[0018] Further, in step S1, the sensitive threshold is preset to ten percent.

[0019] Further, in step S2, the ferrite coil magnetic probe is composed of a ferrite core and an enameled copper wire, the enameled copper wire is wound on the side surface of the ferrite core to form a coil, and the normal magnetic force line passes through the coil during detection.

[0020] Further, in step S5, the value of N is in the range of N≥1.

[0021] Based on the above technical scheme, the IGBT module bonding wire peeling detection method based on the surface magnetic signal has the following advantages:

[0022] 1) No need for electrical direct connection, the present application ingeniously utilizes electromagnetic coupling and electromagnetic induction principle to obtain detection signals without direct electrical connection, thereby eliminating the interference problem between the detection system and the IGBT package module and the potential breakdown risk;

[0023] 2) Non-invasive, relying on the high current change rate during the IGBT module shutdown process, the ferrite coil magnetic probe only needs to be placed on the surface of the package shell to obtain sufficient voltage signals, unlike the traditional detection method based on magnetic signals, which does not need to place the sensor near the current-carrying bonding wire through invasive and destructive means;

[0024] 3) Having the potential for in-situ detection, the present application utilizes the magnetic signals generated during the shutdown process of the IGBT package module as a switching device to realize real-time detection of bonding wire peeling, which has excellent potential for in-situ detection and is widely applicable. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is the mesh division graph of the IGBT geometric model in the embodiment of the present application.

[0026] Figure 2 is the current density distribution graph around the bonding wire of the IGBT geometric model in the embodiment of the present application.

[0027] Figure 3 is the magnetic flux density distribution graph around the bonding wire of the IGBT geometric model in the embodiment of the present application.

[0028] Figure 4 is the surface normal magnetic flux density distribution graph of the IGBT geometric model in the embodiment of the present application.

[0029] Figure 5 is the surface normal magnetic flux density change rate distribution graph of the IGBT geometric model in the embodiment of the present application, Figure 5 (a) in the figure is the bonding wire #1 peeling state, Figure 5 (b) in the figure is the bonding wire #2 peeling state, Figure 5 (c) in the figure is the bonding wire #3 peeling state, Figure 5 (d) in the figure is the bonding wire #4 peeling state.

[0030] Figure 6 is the structure schematic diagram of the double pulse test circuit in the embodiment of the present application.

[0031] Figure 7A schematic diagram of the principle of the IGBT module bonding wire peeling detection based on the surface magnetic signal in the embodiment of the present application.

[0032] Figure 8 A surface induced electromotive force peak rate distribution diagram of the IGBT module to be detected in the single bonding wire peeling state in the embodiment of the present application, Figure 8 In (a), the bonding wire #1 is in the peeling state, Figure 8 In (b), the bonding wire #2 is in the peeling state, Figure 8 In (c), the bonding wire #3 is in the peeling state, Figure 8 In (d), the bonding wire #4 is in the peeling state.

[0033] Figure 9 A bonding wire peeling characterization distribution diagram of the IGBT module to be detected in the embodiment of the present application.

[0034] Figure 10 The induced electromotive force peak rate of each detection point of the IGBT module to be detected in the embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical scheme and beneficial effects of the present application clearer and more apparent, the present application will be further described in detail below with reference to the drawings and embodiments.

[0036] The embodiment proposes an IGBT module bonding wire peeling detection method based on a surface magnetic signal, which realizes the peeling detection of the bonding wire between the IGBT chip and the emitter based on the transient magnetic signal of the surface of the IGBT packaging module excited by the load current in the off process, and specifically includes the following steps:

[0037] S1, modeling and simulating the IGBT module to be detected by the finite element method, and delimiting a rectangular sensitive detection area for bonding wire peeling detection; specifically:

[0038] S11, for the IGBT module to be detected, an IGBT geometric model is established in SOLIDWORKS in a 1:1 reduction manner, including: base copper bottom plate, DBC solder layer, DBC substrate, IGBT chip solder layer, IGBT chip, silicone gel, IGBT packaging shell, etc.; it should be noted that the IGBT geometric model construction process is a known technology in the art, which will not be described here again, but in the present application, the surface magnetic signal detection of the IGBT packaging module needs to be performed, therefore, the modeling process is performed in a 1:1 reduction manner, including the IGBT packaging shell;

[0039] S12. Import the IGBT geometric model into the finite element analysis software COMSOL Multiphysics 6.2, set the material type and parameters of each layer structure, all of which are known a priori; perform finite element mesh generation on the IGBT module to be tested to obtain the mesh generation diagram; mark the side surface of the IGBT geometric model closest to the bonding lead to be tested as the surface to be tested, apply electric and magnetic fields to the IGBT geometric model, and add the excitation condition of the predetermined static load current in the electric field to complete the electromagnetic simulation and obtain the initial normal magnetic flux density distribution of the surface to be tested;

[0040] S13. Maintaining the same simulation conditions as in step S12, set the conductivity of each bonding wire to 10. -12 S / m is used to simulate the stripping of a single bond wire, and the normal magnetic flux density distribution of the surface under test is obtained under each bond wire stripping state. The rate of change of the normal magnetic flux density of the surface under test is calculated under each bond wire stripping state.

[0041] S14. Delineate a rectangular sensitive detection area based on the rate of change of the normal magnetic flux density of the surface to be detected. The rectangular sensitive detection area covers the area where the rate of change of the normal magnetic flux density is greater than 10% under the various bonded wire stripping states.

[0042] In this embodiment, as Figure 1 The image shows the finite element mesh generation diagram of the IGBT geometric model, where I L This is the static load current, specifically 50A; for example... Figure 2 The figure shows the current density distribution of the bonding wires in the IGBT geometric model. The four bonding wires in the IGBT geometric model are labeled #1, #2, #3, and #4. As can be seen from the figure, the current density on the bonding wires is particularly significant. This is because the cross-sectional area of ​​the aluminum bonding wires is much smaller than that of the copper base plate, causing the load current to concentrate on the bonding wires. Figure 3 The figure shows the magnetic flux density distribution of the bonding wires in the IGBT geometric model. According to the Biot-Savart law, the load current excites a flux density of up to several hundred Gauss around the bonding wires (1 Gauss = 10⁻⁻⁴). 4 Tesla's (Tesla's) spatial magnetic field; based on this, such as Figure 4 The figure shows the surface normal magnetic flux density distribution of the IGBT geometric model. The surface normal magnetic flux density of the IGBT geometric model is mainly concentrated on the side surface closest to the bonding wire to be detected, with an intensity reaching tens of Gauss, which is a measurable and significant level. Therefore, in this embodiment, a rectangular area with a length of 4 cm and a width of 3 cm is defined on this side surface as the sensitive detection area. Furthermore, the rate of change of normal magnetic flux density before and after the peeling of bonding wires #1, #2, #3, and #4 is extracted, and the results are as follows... Figure 5As shown, the change rate of the normal magnetic flux density of the peeled bonding wire #1, #2, #3, #4 is shown in sequence as Figure 5 In (a), Figure 5 In (b), Figure 5 In (c), Figure 5 In (d), the results show that the change amplitude of the normal magnetic flux density in the sensitive detection area can reach 10%, that is, the bonding wire peeling can be regionally predicted in the sensitive detection area;

[0043] S2, a double pulse test circuit is built for the IGBT module to be detected, and a ferrite coil magnetic probe is used to scan the rectangular sensitive detection area on the surface of the IGBT module to be detected during the test, and the initial induced electromotive force distribution of the rectangular sensitive detection area is obtained;

[0044] The double pulse test circuit is a key circuit for evaluating the dynamic characteristics of power semiconductor devices, and is widely used in power electronic design and device selection. Its core function is to simulate the transient behavior of switching devices through pulse signals, and measure parameters such as conduction delay and switching loss. For example, Figure 6 As shown, the upper IGBT 1 gate applies a reverse voltage, so that the upper IGBT 1 is always in an off state, and only the freewheeling diode works; the lower IGBT 2 is the IGBT module to be detected, and the falling edge of the first gate pulse (V GE ) simulates the off moment of the IGBT module to be detected in the actual working condition, and the ferrite coil magnetic probe captures the induced electromotive force of the IGBT module to be detected at the off moment;

[0045] The ferrite coil magnetic probe is composed of a ferrite core and an enameled copper wire, and the enameled copper wire is wound on the side surface of the ferrite core to form a coil. In the detection process, the normal magnetic force line passes through the coil. In this embodiment, the diameter of the ferrite core is 2 mm and the length is 7 mm, and the diameter of the enameled copper wire is 0.4 mm. The ferrite coil magnetic probe is installed on the electrically controlled displacement scanning platform, and the rectangular sensitive detection area on the surface of the IGBT module to be detected is scanned by the electrically controlled displacement scanning platform, and the induced electromotive force of each scanning point in the rectangular sensitive detection area is captured, that is, the induced voltage signal. The induced voltage signal is collected by an oscilloscope and transmitted to a host computer for data analysis. In this embodiment, the scanning step is set to 1 mm, and the number of detection points reaches 1200;

[0046] As shown in Figure 7As shown in the principle diagram of the IGBT module bonding wire peeling detection based on the surface magnetic signal in the embodiment, according to the Biot-Savart law, during the off process of the IGBT module to be detected, the load current of the ultra-fast off will inevitably induce a changing magnetic field on the surface of the IGBT module in proportion to the current intensity, based on the Faraday's law of electromagnetic induction, the ferrite coil magnetic probe placed on the surface of the IGBT module will induce a voltage signal in proportion to the normal magnetic flux density rate of change, so it can be known that the voltage signal induced by the ferrite coil magnetic probe is also in proportion to the load current rate of change; considering that the duration of the off process is extremely short (typical value is about 100 nanoseconds), the load current rate of change can easily reach 10 8 A / s, which is the key reason why the present application can induce a strong enough voltage signal on the surface of the IGBT module;

[0047] When the bonding wire peeling occurs, the current density on the bonding wire will be redistributed, causing the flow path of the load current and the current intensity carried by each bonding wire to change, thereby changing the normal magnetic flux density induced on the surface of the IGBT module, and finally reflecting in the change of the voltage signal induced by the ferrite coil magnetic probe, therefore, the bonding wire peeling detection can be realized through the change of the induced electromotive force of the ferrite coil magnetic probe;

[0048] S3, keep the same test process in step S2, cut off each bonding wire of the IGBT module to be detected respectively, measure the induced electromotive force distribution of the rectangular sensitive detection area under the single bonding wire peeling state, and calculate the induced electromotive force rate of change distribution of the rectangular sensitive detection area under the single bonding wire peeling state;

[0049] In the embodiment, as shown in Figure 8 The induced electromotive force peak rate of change distribution diagram of the rectangular sensitive detection area of the IGBT module to be detected under the single peeling state of the bonding wire #1, #2, #3 and #4 is shown in Figure 8 (a) is the bonding wire #1 peeling state, Figure 8 (b) is the bonding wire #2 peeling state, Figure 8 (c) is the bonding wire #3 peeling state, Figure 8 (d) is the bonding wire #4 peeling state; the results show that the induced electromotive force peak rate of change of the rectangular sensitive detection area is significant, and the rectangular sensitive detection area defined in the surface early simulation can realize the bonding wire peeling detection;

[0050] S4, fuse the induced electromotive force rate of change distribution of the rectangular sensitive detection area under the single peeling state of each bonding wire to obtain the bonding wire peeling representation distribution;

[0051] For each scanning point, the minimum value of the rate of change of the electromotive force is taken as the bonding wire peeling characterization value, thereby forming a bonding wire peeling characterization distribution;

[0052] S5, selecting a detection point from the bonding wire peeling characterization distribution and setting an alarm threshold;

[0053] The characterization values in the bonding wire peeling characterization distribution are sorted in descending order, and the scanning points corresponding to the first N characterization values are selected as the detection points, and the number N can be set according to the actual application scenario, N≥1; further, the minimum value of the bonding wire peeling characterization value in all detection points is selected as the alarm threshold;

[0054] Through the above process, the bonding wire peeling characterization distribution can avoid missing detection in the bonding wire peeling detection process, and selecting the discrete scanning points with significant characterization values from the bonding wire peeling characterization distribution as the detection points can ensure the sensitivity of the bonding wire peeling detection; in this embodiment, as shown in Figure 9 Fig. 1 is a bonding wire peeling characterization distribution diagram, from which detection points A, B, C and D are selected, and the characterization values of the detection points are 14.2%, 9.6%, 11.1% and 10.8% respectively, as can be seen from the figure, the detection points A, B, C and D show high detection sensitivity to any bonding wire peeling, and one or more fixed ferrite coil magnetic probes are selected from these detection points;

[0055] S6, fixing the ferrite coil magnetic probe at the detection point of the IGBT module to be detected, and capturing the induced electromotive force in the working process of the IGBT module to be detected in real time, if the rate of change of the induced electromotive force (the rate of change of the induced electromotive force at the current time relative to the induced electromotive force at the previous time) of any ferrite coil magnetic probe exceeds the alarm threshold, it is determined that the bonding wire peeling occurs.

[0056] In summary, the present application proposes an IGBT module bonding wire peeling detection method based on surface magnetic signal, which can complete non-invasive detection without electrical direct connection; more importantly, as Figure 10The peak value change rate of the induced electromotive force of the detection point A, B, C and D in the embodiment shows one-to-one correspondence representation to the bonding wire, for example, the peak value change rate of the induced electromotive force of the detection point A in the peeling state of the bonding wire #1 is much higher than that of other bonding wires, and the peak value change rate of the induced electromotive force of the detection point B in the peeling state of the bonding wire #4 is much higher than that of other bonding wires, the peak value change rate of the induced electromotive force of the detection point C in the peeling state of the bonding wire #3 is much higher than that of other bonding wires, and the peak value change rate of the induced electromotive force of the detection point D in the peeling state of the bonding wire #2 is much higher than that of other bonding wires, so it can be seen that through reasonable selection of the detection point, the IGBT module of the present application can not only realize alarm when the bonding wire peels off, but also has the potential to realize positioning of the peeled bonding wire.

[0057] The above is only a specific embodiment of the present application, any feature disclosed in the specification can be replaced by other equivalent or similar purpose alternative features unless specifically described, and all features disclosed or steps in all methods or processes can be combined in any way except for mutually exclusive features and / or steps.

Claims

1. A method for detecting wire bond peel of an IGBT module based on surface magnetic signal, characterized in that, The method comprises the following steps: S1, modeling and simulating the IGBT module to be detected by the finite element method, marking the side surface closest to the bonding wire to be detected in the geometric model of the IGBT module to be detected as a detection surface, simulating the peeling of a single bonding wire during the simulation process, obtaining the rate of change of the normal magnetic flux density of the detection surface under the condition that each bonding wire is peeled off alone, and delimiting a rectangular sensitive detection area according to the rate of change of the normal magnetic flux density of the detection surface, wherein the rectangular sensitive detection area covers the area where the rate of change of the normal magnetic flux density under the condition that each bonding wire is peeled off alone is greater than a preset sensitive threshold value; S2, building a double-pulse test circuit for the IGBT module to be detected, and performing field scanning on the rectangular sensitive detection area on the surface of the IGBT module to be detected by using a ferrite coil magnetic probe during the test process to obtain the initial induced electromotive force distribution of the rectangular sensitive detection area; S3, keeping the same test process as in S2, cutting off each bonding wire of the IGBT module to be detected, measuring the induced electromotive force distribution of the rectangular sensitive detection area under the condition that a single bonding wire is peeled off, and calculating the induced electromotive force change rate distribution of the rectangular sensitive detection area under the condition that each bonding wire is peeled off alone by subtracting the initial induced electromotive force distribution; S4, fusing the induced electromotive force change rate distribution of the rectangular sensitive detection area under the condition that each bonding wire is peeled off alone, taking the minimum value of the induced electromotive force change rate as the bonding wire peeling representation value for each scanning point, and thus forming a bonding wire peeling representation distribution; S5, sorting the representation values in the bonding wire peeling representation distribution in descending order, selecting the scanning points corresponding to the first N representation values as detection points, and selecting the minimum value of the bonding wire peeling representation value among all the detection points as an alarm threshold value; S6, fixing ferrite coil magnetic probes at the detection points of the IGBT module to be detected, capturing the induced electromotive force in real time during the operation of the IGBT module to be detected, and determining that the bonding wire peeling occurs when the induced electromotive force change rate of any ferrite coil magnetic probe exceeds the alarm threshold value.

2. The surface magnetic signal-based IGBT module bondwire peel detection method of claim 1, wherein, The specific process of step S1 is as follows: S11, for the IGBT module to be detected, an IGBT geometric model is established in SOLIDWORKS; S12, the IGBT geometric model is imported into a finite element analysis software, and the side surface closest to the bonding wire to be detected in the IGBT geometric model is marked as a detection surface; the IGBT module to be detected is first subjected to finite element meshing to obtain a meshing diagram; then an electric field and a magnetic field are applied to the IGBT geometric model, and an excitation condition of a static load current is added to the electric field to complete electromagnetic simulation, and an initial normal magnetic flux density distribution of the detection surface is obtained; S13, keep the same simulation conditions in S12, set the conductivity of each bonding wire to 10 -12 S / m to simulate the peeling of a single bonding wire, and obtain the normal magnetic flux density distribution of the surface to be detected under the peeling state of each bonding wire, and calculate the change rate of the normal magnetic flux density of the surface to be detected under the peeling state of each bonding wire. S14, a rectangular sensitive detection area is delimited according to the rate of change of the normal magnetic flux density of the detection surface, wherein the rectangular sensitive detection area covers the area where the rate of change of the normal magnetic flux density under the condition that each bonding wire is peeled off alone is greater than a preset sensitive threshold value.

3. The surface magnetic signal-based IGBT module bondwire peel detection method of claim 1, wherein, In step S1, the sensitive threshold value is preset to be 10%.

4. The surface magnetic signal-based IGBT module bondwire peel detection method of claim 1, wherein, In step S2, the ferrite coil magnetic probe is composed of a ferrite core and an enameled copper wire, the enameled copper wire is wound on the side surface of the ferrite core to form a coil, and during the detection process, the normal magnetic force line passes through the coil.

5. The surface magnetic signal-based IGBT module bondwire peel detection method of claim 1, wherein, In step S5, the value range of N is N>=1.

Images