Power device failure positioning method and system based on multi-physics field coupling

Abstract

The invention provides a power device failure positioning method and system based on multi-physics field coupling, and the method comprises the steps: applying dynamic electric stress to a tested power device, and synchronously collecting the thermal field data, electric field data and mechanical field data of the tested power device; wherein the thermal field data is used for representing the surface temperature distribution of the to-be-measured power device, the electric field data is used for representing the electrical parameters of the to-be-measured power device, and the mechanical field data is used for representing the internal morphology of the to-be-measured power device; synchronizing the thermal field data, the electric field data and the mechanical field data through timestamps, and aligning in a preset three-dimensional coordinate system to obtain multi-dimensional alignment data; and generating a failure probability distribution diagram according to the multi-dimensional alignment data so as to determine corresponding failure point coordinates. Through multi-physics field data coupling analysis, the problem of failure point fuzzy positioning in a complex packaging structure is solved, and the detection sensitivity under a dynamic working condition is improved.

Description

Technical Field

[0001] This invention relates to the field of semiconductor device testing technology, and in particular to a method and system for locating failures in power devices based on multi-physics coupling. Background Technology

[0002] Traditional failure location techniques (such as thermal imaging, OBIRCH, and EMMI) rely on a single physical field signal (thermal or electrical), making it difficult to distinguish the cumulative effect of failure points in multi-layered packaging structures. Under dynamic operating conditions, the electrothermal coupling effect makes it difficult to capture the transient characteristics of failure points, resulting in low location accuracy; while acoustic emission technology has insufficient sensitivity for microcrack detection and is easily affected by environmental noise. Summary of the Invention

[0003] This invention provides a method and system for locating power device failures based on multi-physical field coupling, in order to solve the technical problems that traditional single physical field signals are difficult to distinguish the superposition effect of failure points and have low positioning accuracy under dynamic operating conditions.

[0004] This invention provides a power device failure location method based on multi-physics coupling, comprising: applying dynamic electrical stress to the power device under test and simultaneously acquiring thermal field data, electric field data, and mechanical field data of the power device under test; wherein, the thermal field data is used to characterize the surface temperature distribution of the power device under test, the electric field data is used to characterize the electrical parameters of the power device under test, and the mechanical field data is used to characterize the internal morphology of the power device under test; synchronizing the thermal field data, the electric field data, and the mechanical field data through timestamps and aligning them in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data; generating a failure probability distribution map based on the multi-dimensional aligned data to determine the coordinates of the corresponding failure point.

[0005] In one embodiment of the present invention, the mechanical field data includes the frequency characteristics and energy distribution of the acoustic emission signal; the electric field data includes the source leakage current and on-state voltage drop of the power device under test; and the thermal field data includes thermal imaging images.

[0006] In one embodiment of the present invention, the step of synchronizing the thermal field data, the electric field data and the mechanical field data through timestamps includes: acquiring the time-domain waveforms of each electrical parameter in the electric field data and the acoustic wave image corresponding to the acoustic emission signal; and synchronizing the thermal imaging image, the time-domain waveform and the acoustic wave image through timestamps.

[0007] In one embodiment of the present invention, the step of generating a failure probability distribution map based on the multidimensional alignment data includes: extracting features from the multidimensional alignment data using a pre-constructed thermo-electro-mechanical coupling model to obtain a thermo-electric correlation coefficient and a time delay between acoustic emission time and temperature change; identifying thermal stress-dominated crack regions based on the time delay; and performing weighted fusion of the thermo-electric correlation coefficient and the crack regions to obtain the failure probability distribution map, wherein the failure probability distribution map is marked with the coordinates of candidate failure points.

[0008] In one embodiment of the present invention, after obtaining the failure probability distribution map, the method further includes: performing profile verification on the candidate failure points in the failure probability distribution map using a micro-area focused ion beam to determine the failure type corresponding to each candidate failure point; and adjusting the weight parameters of the weighted fusion according to the matching degree between the failure type and the corresponding candidate failure point's annotation information in the failure probability distribution map.

[0009] In one embodiment of the present invention, the step of alignment in a preset three-dimensional coordinate system includes: selecting a surface as a reference surface in the preset three-dimensional coordinate system; calculating the three-dimensional equipotential height of the thermal imaging image, the time-domain waveform, and the acoustic image relative to the reference surface, thereby completing image fusion and obtaining the multidimensional alignment data.

[0010] This invention also provides a system for executing the aforementioned multi-physics coupling power device failure location method, comprising: a multi-channel signal generator for applying dynamic electrical stress to the power device under test; an infrared thermal imager for acquiring thermal field data of the power device under test; an acoustic emission sensor array for acquiring mechanical field data of the power device under test; a data acquisition card for acquiring electric field data of the power device under test; a processing module for synchronizing the thermal field data, the electric field data, and the mechanical field data through timestamps and aligning them in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data; and generating a failure probability distribution map based on the multi-dimensional aligned data to determine the coordinates of the corresponding failure point.

[0011] In one embodiment of the present invention, the processing module further includes a multi-physics data synchronization engine, which is configured to acquire the time-domain waveforms of each electrical parameter in the electric field data; and to synchronize the thermal imaging image corresponding to the thermal field data, the time-domain waveforms, and the acoustic emission signals corresponding to the mechanical field data through timestamps.

[0012] In one embodiment of the present invention, the processing module further includes a coupling feature analysis algorithm library, which stores a pre-established thermo-electro-mechanical coupling model. The beneficial effects of the present invention are as follows: The present invention proposes a method and system for locating power device failures based on multi-physics coupling. Through multi-physics data coupling analysis, it solves the problem of fuzzy location of failure points in complex packaging structures and improves the detection sensitivity under dynamic operating conditions. Attached Figure Description

[0013] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0014] In the attached diagram: Figure 1 This is a schematic diagram of the overall architecture of a power device failure location system based on multi-physics coupling in one embodiment of the present invention; Figure 2 This is a flowchart illustrating a power device failure location method based on multi-physics coupling provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the result of multiphysics data fusion in one embodiment of the present invention.

[0015] The attached figures are labeled as follows: 01-Base; 02-Stage; 03-Signal transmission channel; 04-Infrared thermal imager; 05-Support rod; 06-Clamping part; 07-Processing module; 08-Power supply; 09-Acoustic transducer. Detailed Implementation

[0016] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0017] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. The drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0018] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0019] Please see Figure 1 , Figure 1 This is a schematic diagram of the overall architecture of a power device failure location system based on multi-physics coupling according to an embodiment of the present invention. The system includes a base 01, a stage 02, a multi-channel signal generator, an infrared thermal imager 04, a data acquisition card, an acoustic emission sensor array, and a processing module 07. The stage 02 is used to support the power device under test and is placed on the base 01. The multi-channel signal generator is used to apply dynamic electrical stress to the power device under test; the infrared thermal imager 04 is used to collect thermal field data of the power device under test; the acoustic emission sensor array is used to collect mechanical field data of the power device under test; the data acquisition card is used to collect electric field data of the power device under test; the processing module 07 is used to synchronize the thermal field data, the electric field data, and the mechanical field data through timestamps and align them in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data; and to generate a failure probability distribution map based on the multi-dimensional aligned data to determine the coordinates of the corresponding failure point. Specifically, a slot can be cut into the stage 02 as a mounting position for placing the power device under test. The mounting position includes probes for connecting to the pins of the power device under test (DUT). The multi-channel signal generator includes multiple signal output channels 03 for providing dynamic electrical stress excitation. Each channel can transmit a different signal. The multiple signal transmission channels 03 of the multi-channel signal generator can be positioned to the side of the stage 02, and each signal transmission channel 03 can protrude above the plane of the stage 02 facing away from the substrate. Each signal transmission channel 03 can be connected to the corresponding pin of the DUT via leads; the specific pins connected can be set and adjusted according to actual testing requirements. Figure 2 Only four signal transmission channels 03 are shown in the diagram. These four channels are distributed around the stage and extend leads to connect to the power device under test (DUT) placed on the stage. A first interface is provided on the base 01, through which the multiple signal transmission channels 03 establish an electrical connection with the external signal generator. A cavity can be provided inside the base 01, and an acoustic emission sensor array is placed inside the cavity, positioned directly below the mounting location of the DUT. The acoustic emission sensor array can acquire acoustic wave images of the DUT, which reflect the internal morphology of the DUT.

[0020] In one embodiment, a second interface may be provided on the base 01, through which the acoustic emission sensor array is electrically connected to an external acoustic transducer 09. Mechanical field data is acquired through the acoustic emission sensor array. Specifically, the acoustic emission sensor array may employ a broadband acoustic emission sensor (20kHz-50MHz). The broadband acoustic emission sensor displays the internal morphology of the power device under test as a grayscale image based on the amplitude of the sound wave in the time domain; this grayscale image is the acoustic wave image. The acoustic wave image is part of the mechanical field data, which may also include information such as the frequency characteristics and energy distribution of the acoustic emission signal.

[0021] In one embodiment, the base 01 may further include a power interface, through which an external power supply 08 establishes an electrical connection with the power device under test. Specifically, the power interface is connected to the probe in the mounting position.

[0022] In one embodiment, a support rod 05 may be provided on the base 01, and a clamping part 06 is connected to the support rod 05. The clamping part 06 can slide along the axial direction of the support rod 05 or be fixed to a predetermined position on the support rod 05. The clamping part 06 extends radially outward along the support rod 05. The distal end of the clamping part 06 is used to fix an infrared thermal imager, so that the infrared thermal imager is facing the area where the mounting position is located on the stage. When the power device under test is placed on the mounting position, thermal field data can be collected by the infrared thermal imager above, and the thermal field data may include infrared thermal imaging images.

[0023] In one embodiment, a data acquisition card is also installed inside the base 01. The data acquisition card has a synchronous sampling rate of ≥1MS / s. The data acquisition card can establish an electrical connection with the power device under test through the probes of the mounting position. Electric field data is acquired through the data acquisition card. Specifically, key electrical parameters such as the source leakage current (Idss) and on-state voltage drop (Vds(on)) of the power device under test can be acquired in real time.

[0024] In one embodiment, the processing module 07 is electrically connected to the infrared thermal imager 04, the data acquisition card, the acoustic emission sensor array, and the multi-channel signal generator, respectively. It can receive the acquired data, analyze the acquired data, and output data such as the coordinates of the failure point.

[0025] Please see Figure 2 , Figure 2 The following is a flowchart illustrating a power device failure location method based on multi-physics coupling according to an embodiment of the present invention. The method includes the following steps: Step S200: Apply dynamic electrical stress to the power device under test and simultaneously collect thermal field data, electric field data and mechanical field data of the power device under test; wherein, the thermal field data is used to characterize the surface temperature distribution of the power device under test, the electric field data is used to characterize the electrical parameters of the power device under test, and the mechanical field data is used to characterize the internal morphology of the power device under test.

[0026] In one embodiment, the dynamic electrical stress can be a pulsed current, voltage ramp, etc., and can be connected to the power device under test through a multi-channel signal generator to provide the corresponding dynamic electrical stress excitation. The specific number of channels of the signal generator can be designed and adjusted according to actual test requirements, and is not limited here.

[0027] In one embodiment, the temperature distribution on the surface of the power device under test can be acquired using a high-resolution infrared thermal imager (sampling rate ≥100Hz) to obtain an infrared thermal image. This high-resolution infrared thermal imager can integrate a limiting optical lens to magnify the image acquisition area and obtain a clearer infrared thermal image. The infrared thermal imager can be installed directly facing the area where the power device under test is placed.

[0028] Step S210: Synchronize the thermal field data, the electric field data, and the mechanical field data using timestamps, and align them in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data.

[0029] In one embodiment, the step of synchronizing the thermal field data, electric field data, and mechanical field data using timestamps includes: acquiring the time-domain waveforms of each electrical parameter in the electric field data and the acoustic wave image corresponding to the acoustic emission signal; synchronizing the thermal imaging image, the time-domain waveform, and the acoustic wave image using timestamps. A high-speed data acquisition card is used to digitize the original signal, low-pass filtering is used to eliminate high-frequency noise, and then the fundamental and harmonic components are extracted using the FFT (Fast Fourier Transform) algorithm, ultimately generating a time-domain waveform diagram containing parameters such as voltage peak value, frequency characteristics, and phase difference. An acoustic emission sensor array captures ultrasonic signals generated by partial discharge or mechanical vibration. After band-pass filtering and signal amplification, a beamforming algorithm is used to spatially locate the multi-channel data, generating acoustic wave images (such as C-scan or B-scan images) reflecting the location and intensity of the sound source. A high-precision GPS clock module is deployed in each data acquisition system, embedding a UTC timestamp in the data frame header. For electric field data, synchronization can be achieved using the IEEE 1588 Precision Time Protocol (PTP); acoustic emission signals are aligned with the clock of the data acquisition card via a hardware trigger signal; and the thermal imager is synchronized with the processing module via an external trigger interface.

[0030] In one embodiment, the step of aligning in a preset three-dimensional coordinate system includes: selecting a surface as a reference surface in the preset three-dimensional coordinate system; calculating the three-dimensional equipotential height of the thermal imaging image, the time-domain waveform, and the acoustic image relative to the reference surface, thereby completing image fusion and obtaining the multidimensional alignment data.

[0031] Step S220: Generate a failure probability distribution map based on the multidimensional alignment data to determine the coordinates of the corresponding failure point.

[0032] In one embodiment, a pre-constructed thermo-electro-mechanical coupling model is used to extract features from the multidimensional aligned data to obtain a thermo-electric correlation coefficient, which yields the temporal delay between acoustic emission time and temperature change. Based on this temporal delay, thermal stress-dominated crack regions are identified. A weighted fusion of the thermo-electric correlation coefficient and the crack regions is then performed to obtain the failure probability distribution map, wherein the failure probability distribution map is marked with the coordinates of candidate failure points. Please refer to [link to relevant documentation]. Figure 3 , Figure 3 This is a schematic diagram illustrating the results of multiphysics data fusion in one embodiment of the present invention. A three-dimensional finite element model incorporating heat conduction, electric field distribution, and mechanical stress is established. The model parameters are calibrated using experimental data (such as infrared thermometry, partial discharge detection, and strain gauge measurements). A particle swarm optimization (PSO) algorithm is used to minimize the error between the model's predicted and measured values, ensuring that the accuracy of the coupled model is greater than the set target accuracy (e.g., 95%). The correlation coefficient between local temperature rise (ΔT) and leakage current change rate (dIdss / dt) is calculated. The time delay between acoustic emission events and temperature changes is analyzed to identify thermal stress-dominated crack regions. An adaptive weighted fusion algorithm is used to output a failure probability distribution map and mark the coordinates of candidate failure points.

[0033] In one embodiment, after obtaining the failure probability distribution map, the method further includes: performing profile verification on candidate failure points in the failure probability distribution map using a micro-area focused ion beam (FIB) to determine the failure type corresponding to each candidate failure point; and adjusting the weight parameters of the weighted fusion based on the matching degree between the failure type and the corresponding candidate failure point's annotation information in the failure probability distribution map. Specifically, micro-area focused ion beam (FIB) profile verification is performed on the candidate failure points, and the model weight parameters are optimized inversely based on the actual failure type (such as gate oxide breakdown or solder voids).

[0034] Specifically, for the short-circuit failure analysis of the 1200V SiC MOSFET, a pulse voltage of 800V, a pulse width of 10μs, and a repetition frequency of 1kHz were applied; an infrared thermal imager detected an instantaneous temperature rise of ΔT = 85℃ in the central region of the chip; an acoustic emission sensor array captured a stress wave signal with an inverted amplitude characteristic (corresponding to microcrack propagation); the leakage current showed an abnormal spike (amplitude > 1mA) at the falling edge of the pulse. Through model analysis, it was determined that the failure point was located at the edge of the cell region, and FIB verification confirmed it as local breakdown of the gate oxide.

[0035] The technical solution of this invention reduces the failure point location error from >10μm to ≤3μm by multi-field coupling analysis, supports failure capture under transient conditions (such as short-circuit impact), and has a time resolution of 1μs; it is applicable to various power device types such as TO-247 and module packages.

[0036] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for locating failures in power devices based on multiphysics coupling, characterized in that, include: Dynamic electrical stress is applied to the power device under test, and thermal field data, electric field data and mechanical field data of the power device under test are collected simultaneously; wherein, the thermal field data is used to characterize the surface temperature distribution of the power device under test, the electric field data is used to characterize the electrical parameters of the power device under test, and the mechanical field data is used to characterize the internal morphology of the power device under test. The thermal field data, electric field data, and mechanical field data are synchronized using timestamps and aligned in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data. A failure probability distribution map is generated based on the multidimensional alignment data to determine the coordinates of the corresponding failure points.

2. The power device failure location method based on multiphysics coupling according to claim 1, characterized in that, The mechanical field data includes the frequency characteristics and energy distribution of the acoustic emission signal; the electric field data includes the source leakage current and on-state voltage drop of the power device under test; and the thermal field data includes thermal imaging images.

3. The power device failure location method based on multiphysics coupling according to claim 2, characterized in that, The step of synchronizing the thermal field data, the electric field data, and the mechanical field data using timestamps includes: Obtain the time-domain waveforms of each electrical parameter in the electric field data and the acoustic wave image corresponding to the acoustic emission signal; The thermal imaging image, the time-domain waveform, and the acoustic image are synchronized using timestamps.

4. The power device failure location method based on multiphysics coupling according to claim 1, characterized in that, The steps for generating a failure probability distribution map based on the multidimensional alignment data include: Feature extraction is performed on the multidimensional aligned data using a pre-constructed thermo-electro-mechanical coupling model to obtain the thermo-electro-correlation coefficient, thus obtaining the temporal delay between acoustic emission time and temperature change. Based on the aforementioned timing delay, thermal stress-dominated crack regions are identified; The failure probability distribution map is obtained by weighting and fusing the thermo-electric correlation coefficient with the crack region, wherein the coordinates of candidate failure points are marked in the failure probability distribution map.

5. The power device failure location method based on multiphysics coupling according to claim 4, characterized in that, After obtaining the failure probability distribution map, the method further includes: The candidate failure points in the failure probability distribution map are cross-sectionally verified by micro-area focused ion beam to determine the failure type corresponding to each candidate failure point. The weight parameters of the weighted fusion are adjusted according to the degree of matching between the failure type and the corresponding candidate failure point in the failure probability distribution map.

6. The power device failure location method based on multiphysics coupling according to claim 4, characterized in that, The steps for alignment in a preset 3D coordinate system include: Select a surface as a reference surface in the preset three-dimensional coordinate system; The three-dimensional equipotential heights of the thermal imaging image, the time-domain waveform, and the acoustic image relative to the reference surface are calculated, and then image fusion is performed to obtain the multidimensional alignment data.

7. A system for implementing the power device failure location method based on multiphysics coupling as described in any one of claims 1-6, characterized in that, include: A multi-channel signal generator is used to apply dynamic electrical stress to the power device under test; An infrared thermal imager is used to collect thermal field data of the power device under test. An acoustic emission sensor array is used to collect mechanical field data of the power device under test; A data acquisition card is used to acquire the electric field data of the power device under test. The processing module is used to synchronize the thermal field data, the electric field data, and the mechanical field data through timestamps, and align them in a preset three-dimensional coordinate system to obtain multi-dimensional aligned data; A failure probability distribution map is generated based on the multidimensional alignment data to determine the coordinates of the corresponding failure points.

8. The system according to claim 7, characterized in that, The processing module further includes a multi-physics data synchronization engine, which is configured to acquire the time-domain waveforms of each electrical parameter in the electric field data; and to synchronize the thermal imaging image corresponding to the thermal field data, the time-domain waveforms, and the acoustic emission signals corresponding to the mechanical field data through timestamps.

9. The system according to claim 7, characterized in that, The processing module also includes a coupling feature analysis algorithm library, which stores pre-established thermo-electro-mechanical coupling models.

Images