A SiC MOSFET damage accumulation pattern recognition method based on feature point trajectories

Abstract

The application discloses a SiC MOSFET damage accumulation mode recognition method based on feature point trajectories and relates to the technical field of power semiconductor device reliability. The damage accumulation law is defined as three basic modes; the SiC MOSFET constant stress accelerated life test and two-stage stress accelerated life test are designed and carried out, the reliable life of the SiC MOSFET under different test conditions is calculated; the feature point coordinates corresponding to the two-stage stress accelerated life test are calculated based on the reliable life, and the damage accumulation mode of the SiC MOSFET is recognized according to the feature point trajectory inspection. The three basic modes of damage accumulation are defined, the accelerated life test is designed and carried out, and the feature point trajectory inspection is carried out, so that the damage accumulation mode of the SiC MOSFET can be accurately recognized through a simple test mode.

Description

Technical Field

[0001] This invention relates to the field of power semiconductor device reliability technology, specifically a method for identifying damage accumulation patterns in SiCMOSFETs based on feature point trajectories. Background Technology

[0002] As a new generation of semiconductor devices, SiC MOSFETs have greatly promoted the development of power electronic converters due to their superior performance. However, because SiC's Young's modulus is three times that of Si, the packaging reliability problem of SiC MOSFETs is more serious. Against this background, lifetime prediction for SiC MOSFETs is extremely important.

[0003] The foundation for lifetime prediction of SiC MOSFETs under dynamic mission profiles lies in identifying the damage accumulation modes of SiC MOSFETs. However, most current research neglects this aspect, resulting in low lifetime prediction accuracy. Therefore, developing a method capable of identifying the damage accumulation modes of SiC MOSFETs is imperative. Summary of the Invention

[0004] To address the shortcomings of the prior art, this invention provides a method for identifying damage accumulation modes of SiC MOSFETs based on feature point trajectories. It defines three basic damage accumulation modes, designs and conducts accelerated life tests and feature point trajectory verification, and achieves accurate identification of damage accumulation modes of SiC MOSFETs through a simple experimental approach.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories, comprising the following steps:

[0006] Step 1: Define the damage accumulation law into three basic modes, namely Mode 1, Mode 2 and Mode 3. Mode 1 is that the damage rate is determined only by the current stress level. Mode 2 is that the damage rate is determined by the current stress level and the damage value. Mode 3 is that the damage rate is determined by the current stress level, the damage value and the historical stress conditions.

[0007] Step 2: Design and conduct constant stress accelerated life test and two-stage stress accelerated life test of SiC MOSFET, and calculate the reliable life of SiC MOSFET under different test conditions;

[0008] Step 3: Calculate the coordinates of the feature points corresponding to the two-stage stress-accelerated life test based on the reliable life calculation, and identify the damage accumulation mode of the SiC MOSFET based on the feature point trajectory.

[0009] Furthermore, in step one, when the damage accumulation is mode 1, the damage accumulates linearly and the slope of the damage accumulation curve changes after the stress switch; when the damage accumulation is mode 2, the damage accumulates nonlinearly and still accumulates along the damage accumulation curve corresponding to constant stress after the stress switch; when the damage accumulation is mode 3, the damage accumulation curve deviates from the damage accumulation curve corresponding to constant stress after the stress switch.

[0010] Furthermore, step two specifically includes:

[0011] S2.1, Conduct two sets of junction temperature swings, ΔT j1 and ΔT j2 The SiC MOSFET constant stress accelerated life test was conducted, and the reliable lifetime L1(x%) and L2(x%) of the SiC MOSFET under two sets of constant stress test conditions were calculated using the Weibull distribution, where x% represents the reliability.

[0012] S2.2 Conduct two sets of two-stage stress-accelerated life tests, where the junction temperature fluctuation in the first stage is ΔT. j1 The junction temperature swing in the second stage is ΔT j2 The number of cycles in the first stage corresponding to the two sets of two-stage stress tests are n respectively. 1a and n 1b The reliable lifetime n of SiC MOSFETs under two sets of two-stage stress test conditions was calculated using the Weibull distribution, corresponding to the number of cycles in the second stage. 2a (x%) and n 2b (x%).

[0013] Furthermore, in S2.1, ΔT j1 and ΔT j2 The selection should ensure that the median lifetimes L1(50%) and L2(50%) of the SiCMOSFET under the two sets of constant stress test conditions differ by at least 5 times.

[0014] Furthermore, in S2.2, n 1a and n 1b The selection should ensure that 0.2L1(50%)≤n 1a ≤0.3L1(50%), 0.7L1(50%)≤n 1b ≤0.8L1(50%).

[0015] Furthermore, step three specifically includes:

[0016] S3.1 Calculate the coordinates of the characteristic points corresponding to the two-stage stress-accelerated life test. The calculation formula is as follows:

[0017]

[0018] In the formula, ξ a (x a (x%),y a (x%) and ξ b (x b (x%),y b (x%) represents the coordinates of the feature point, and x% represents the reliability.

[0019] S3.2. Perform equidistant sampling on x%, denoted as M, and calculate the point set. and The least squares fit is performed on the data in A∪B using the form of the straight line y=kx to obtain k, and the consistency between the feature point trajectory and the straight line y=kx is judged based on the mean absolute error percentage (MAPE). If the MAPE is greater than the critical value, the damage accumulation mode of the SiC MOSFET is considered to be mode 3; otherwise, S3.3 is executed.

[0020] S3.3 Calculate the mean absolute error percentage (MAPE) between the data in A∪B and the straight line y=x. If it is greater than the critical value, the damage accumulation mode of the SiC MOSFET is considered to be mode 2; otherwise, it is mode 1.

[0021] Furthermore, in S3.2 and S3.3, the critical value is 20%.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: In view of the problem that the damage accumulation law of SiC MOSFET is unclear, the present invention defines three basic modes of damage accumulation, designs and carries out accelerated life test and feature point trajectory inspection, and achieves accurate identification of SiC MOSFET damage accumulation mode through simple test method, providing an effective solution for revealing the law of SiC MOSFET damage accumulation, and can further serve the development of new cumulative damage models, improve the accuracy of lifetime prediction, and can also be further extended to other types of devices. Attached Figure Description

[0023] Figure 1 This is a flowchart of the method of the present invention;

[0024] Figure 2 This is a schematic diagram of damage accumulation curves under different modes in the method of the present invention;

[0025] Figure 3 These are schematic diagrams of feature point trajectories under different modes in the method of this invention;

[0026] Figure 4 This is a feature point trajectory diagram corresponding to the two-stage stress-accelerated life test of the SiC MOSFET in the embodiment. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0028] like Figures 1-3 As shown, a method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories is presented, and its process combines... Figure 1 As shown, it includes the following steps:

[0029] Step 1: Define three damage accumulation modes;

[0030] The damage accumulation law is defined as three basic modes, namely mode 1, mode 2 and mode 3. Mode 1 is that the damage rate is determined only by the current stress level. Mode 2 is that the damage rate is determined by the current stress level and the damage value. Mode 3 is that the damage rate is determined by the current stress level, the damage value and the historical stress conditions.

[0031] Combination Figure 2 As shown, when damage accumulation is mode 1, the damage accumulates linearly and the slope of the damage accumulation curve changes after stress switching. When damage accumulation is mode 2, the damage accumulates nonlinearly and continues to accumulate along the damage accumulation curve corresponding to constant stress after stress switching. When damage accumulation is mode 3, the damage accumulation curve deviates from the damage accumulation curve corresponding to constant stress after stress switching.

[0032] Step 2: Design and conduct constant stress accelerated life test and two-stage stress accelerated life test of SiC MOSFET, and calculate the reliable life of SiC MOSFET under different test conditions;

[0033] S2.1, Conduct two sets of junction temperature swings, ΔT j1 and ΔT j2 The SiC MOSFET constant stress accelerated life test was conducted, and the reliable lifetime L1(x%) and L2(x%) of the SiC MOSFET under two sets of constant stress test conditions were calculated using the Weibull distribution, where x% represents the reliability.

[0034] Preferably, ΔT j1 and ΔT j2 The selection should ensure that the median lifetimes L1(50%) and L2(50%) of the SiC MOSFET under the two sets of constant stress test conditions differ by at least 5 times.

[0035] S2.2 Conduct two sets of two-stage stress-accelerated life tests, where the junction temperature fluctuation in the first stage is ΔT. j1 The junction temperature swing in the second stage is ΔT j2 The number of cycles in the first stage corresponding to the two sets of two-stage stress tests are n respectively. 1a and n 1b The reliable lifetime n of SiC MOSFETs under two sets of two-stage stress test conditions was calculated using the Weibull distribution, corresponding to the number of cycles in the second stage. 2a (x%) and n 2b (x%), where x% represents reliability;

[0036] Preferred, n 1a and n 1b The selection should ensure that 0.2L1(50%)≤n 1a ≤0.3L1(50%), 0.7L1(50%)≤n 1b ≤0.8L1(50%).

[0037] Step 3: Calculate the coordinates of the feature points corresponding to the two-stage stress-accelerated life test based on the reliable life calculation, and identify the damage accumulation mode of the SiC MOSFET based on the feature point trajectory.

[0038] S3.1 Calculate the coordinates of the characteristic points corresponding to the two-stage stress-accelerated life test. The calculation formula is as follows:

[0039]

[0040] In the formula, ξ a (x a (x%),y a (x%) and ξ b (x b (x%),y b (x%) represents the coordinates of the feature point, and x% represents the reliability.

[0041] Combination Figure 3 As shown, when the reliability x% changes, the feature point trajectories corresponding to the three basic modes have the following characteristics:

[0042] Pattern 1: Feature points move near a straight line with a slope of 1 and an intercept of 0;

[0043] Mode 2: The feature point moves near a straight line with a slope of k and an intercept of 0, where k is related to the stress level of the two stages but not to the number of cycles corresponding to the first stage, and k ≠ 1;

[0044] Mode 3: The trajectory of the feature point is related to the stress level of the two stages and the number of cycles corresponding to the first stage.

[0045] S3.2. Based on MAPE, determine the consistency between the feature point trajectory and the straight line y=kx. Specifically, perform equal-interval sampling on x%, denoted as M, and calculate the point set. and We use the form y=kx to perform least squares fitting on the data in A∪B to obtain k, and use the following formula to calculate the mean absolute error percentage (MAPE):

[0046]

[0047] In the formula, This represents the actual value, that is, the data in A∪B. This represents the predicted value, i.e., the data on the fitted straight line y=kx, where n represents the total number of data points.

[0048] If MAPE is greater than the critical value, the damage accumulation mode of the SiC MOSFET is considered to be mode 3; otherwise, S3.3 is executed.

[0049] S3.3. Based on MAPE, determine the consistency between the feature point trajectory and the straight line y=x. Specifically, calculate the MAPE of the data in A∪B and the straight line y=x. If it is greater than the critical value, the damage accumulation mode of SiC MOSFET is considered to be mode 2; otherwise, it is mode 1.

[0050] Example

[0051] This embodiment uses a certain type of SiC MOSFET as the object to identify damage accumulation mode, designs and conducts accelerated life test, and the specific settings are shown in Table 1. The number of samples in each test group is 8, and the failure criterion of SiC MOSFET is an increase of 20% in conduction voltage.

[0052] Table 1 Accelerated Life Test Setup

[0053]

[0054] The scale parameter η and shape parameter β of the Weibull distribution corresponding to each accelerated life test were calculated using least squares fitting. The fitting results are shown in Table 2.

[0055] Table 2. Fitting results of Weibull distribution

[0056]

[0057] Combination Figure 4 As shown in Table 2, the feature point coordinates ξ are calculated based on the fitting results. a (x a (x%),y a (x%) and ξ b (x b(x%),y b (x%), and let M={0.01:0.01:0.99}, to obtain point sets A and B. Using the form of a straight line y=kx, perform least squares fitting on the data in A∪B, obtaining k=1.1959, and calculate MAPE, which is MAPE=27.89%>20%. Therefore, the damage accumulation mode of this SiCMOSFET is mode 3, and the damage rate is determined by the current stress level, damage value, and historical stress conditions.

[0058] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0059] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories, characterized in that: Includes the following steps: Step 1: Define the damage accumulation law into three basic modes, namely Mode 1, Mode 2 and Mode 3. Mode 1 is that the damage rate is determined only by the current stress level. Mode 2 is that the damage rate is determined by the current stress level and the damage value. Mode 3 is that the damage rate is determined by the current stress level, the damage value and the historical stress conditions. Step 2: Design and conduct constant stress accelerated life tests and two-stage stress accelerated life tests on SiC MOSFETs, and calculate the reliable lifetime of SiC MOSFETs under different test conditions, specifically including: S2.1, Conduct two sets of junction temperature swings, ΔT j1 and ΔT j2 The SiC MOSFET constant stress accelerated life test was conducted, and the reliable lifetime L1(x%) and L2(x%) of the SiC MOSFET under two sets of constant stress test conditions were calculated using the Weibull distribution, where x% represents the reliability. S2.2 Conduct two sets of two-stage stress-accelerated life tests, where the junction temperature fluctuation in the first stage is ΔT. j1 The junction temperature swing in the second stage is ΔT j2 The number of cycles in the first stage corresponding to the two sets of two-stage stress tests are n respectively. 1a and n 1b The reliable lifetime n of SiC MOSFETs under two sets of two-stage stress test conditions was calculated using the Weibull distribution, corresponding to the number of cycles in the second stage. 2a (x%) and n 2b (x%); Step 3: Calculate the coordinates of characteristic points corresponding to the two-stage stress-accelerated life test based on reliable lifetime calculations, and identify the damage accumulation mode of the SiC MOSFET based on the characteristic point trajectory. Specifically, this includes: S3.1 Calculate the coordinates of the characteristic points corresponding to the two-stage stress-accelerated life test. The calculation formula is as follows: In the formula, ξ a (x a (x%),y a (x%) and ξ b (x b (x%),y b (x%) represents the coordinates of the feature point, and x% represents the reliability. S3.

2. Perform equidistant sampling on x%, denoted as M, and calculate the point set. and The least squares fit is performed on the data in A∪B using the form of the straight line y=kx to obtain k, and the consistency between the feature point trajectory and the straight line y=kx is judged based on the mean absolute error percentage (MAPE). If the MAPE is greater than the critical value, the damage accumulation mode of the SiC MOSFET is considered to be mode 3; otherwise, S3.3 is executed. S3.3 Calculate the mean absolute error percentage (MAPE) between the data in A∪B and the straight line y=x. If it is greater than the critical value, the damage accumulation mode of the SiC MOSFET is considered to be mode 2; otherwise, it is mode 1.

2. The method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories according to claim 1, characterized in that: In step one, when the damage accumulation is mode 1, the damage accumulates linearly and the slope of the damage accumulation curve changes after stress switching. When the damage accumulation is in mode 2, the damage accumulates nonlinearly and continues to accumulate along the damage accumulation curve corresponding to constant stress after stress switching. When the damage accumulation is in mode 3, the damage accumulation curve after stress switching deviates from the damage accumulation curve corresponding to constant stress.

3. The method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories according to claim 1, characterized in that: In S2.1, ΔT j1 and ΔT j2 The selection should ensure that the median lifetimes L1(50%) and L2(50%) of the SiC MOSFET under the two sets of constant stress test conditions differ by at least 5 times.

4. The method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories according to claim 1, characterized in that: In S2.2, n 1a and n 1b The selection should ensure that 0.2L1(50%)≤n 1a ≤0.3L1(50%), 0.7L1(50%)≤n 1b ≤0.8L1(50%).

5. The method for identifying damage accumulation patterns in SiC MOSFETs based on feature point trajectories according to claim 1, characterized in that: In S3.2 and S3.3, the critical value is 20%.

Images