Threshold voltage degradation analysis methods, apparatus and computer equipment for power devices

Abstract

This application relates to the field of semiconductor technology, and in particular to a method, apparatus, and computer device for threshold voltage degradation analysis of power devices. The method includes: acquiring the threshold voltage drift and number of short circuits for each of multiple power devices with inconsistent initial case temperatures during multiple short-circuit tests; performing a fitting operation using a threshold voltage degradation model based on the threshold voltage drift and number of short circuits for each power device in each short-circuit test to obtain first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each initial case temperature; calculating second specific values ​​of the model parameters in the threshold voltage degradation model based on the first specific values; and obtaining the threshold voltage drift of the device under analysis using the threshold voltage degradation model based on the second specific values, the initial case temperature of the device under analysis, and the number of short circuits, where the device under analysis and the power devices are of the same model. This method can simultaneously consider the influence of the number of short circuits and the initial case temperature of the power device on the threshold voltage.

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, and in particular to a method, apparatus and computer device for threshold voltage degradation analysis of power devices. Background Technology

[0002] Semiconductor power devices are key electronic components used for controlling and converting electrical energy, and are widely used in power management, motor drives, new energy power generation, electric vehicles, and rail transportation. However, in high-frequency, high-power-density scenarios, short circuits frequently occur due to mis-conduction in semiconductor power devices. For example, SiC MOSFETs are widely used in high-frequency, high-power-density applications due to their wide bandgap, high thermal conductivity, and high electron mobility. However, under high voltage and high power density conditions, when a SiC MOSFET experiences a short circuit, it is simultaneously subjected to high temperature, high voltage, and high current, which can damage the device. Furthermore, after prolonged exposure to high temperature and high voltage, defects in the gate oxide trap charges, causing threshold voltage drift and shortening the device's lifespan. Predicting the lifespan of semiconductor power devices after multiple short circuits would be helpful in the design of power electronic systems.

[0003] Currently, existing models for dynamic trapping of trapped charges require solving multiple semiconductor physical differential equations simultaneously, necessitating an iterative approach to obtain modeling and simulation results. This modeling and simulation method is time-consuming and contains significant uncertainties, making it difficult to apply to multiple short-circuit conditions. Furthermore, existing analytical models only consider the impact of stress time on threshold voltage degradation. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, and computer device for threshold voltage degradation analysis of power devices to address the aforementioned technical problems. This method can simultaneously consider the effects of the number of short circuits and the initial case temperature of the power device on the threshold voltage.

[0005] A method for threshold voltage degradation analysis of power devices, the method comprising:

[0006] S1. Obtain the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests.

[0007] S2. Based on the threshold voltage drift and the number of short circuits of each power device in each short circuit test, a threshold voltage degradation model is used for fitting calculation to obtain the first specific value of the intermediate parameters of the threshold voltage degradation model at each initial case temperature.

[0008] S3. Based on the first specific value, calculate the second specific value of the model parameters in the threshold voltage degradation model; the threshold voltage degradation model is... , , K, A1, A2 and All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and n SC For the number of short circuits, This represents the threshold voltage drift.

[0009] S4. Based on the second specific value, the initial case temperature of the device to be analyzed, and the number of short circuits, the threshold voltage drift of the device to be analyzed is obtained through the threshold voltage degradation model. The device to be analyzed and the power device have the same model, and the initial case temperature of the device to be analyzed is arbitrary.

[0010] In this application, by using and By determining the specific values ​​of the intermediate parameters k1 and k2 in the threshold voltage degradation model, the influence of the initial case temperature T of the device under analysis on the threshold voltage can be considered. Since the threshold voltage degradation model is related to the number of short circuits of the device under analysis, this application can simultaneously consider the influence of the number of short circuits and the initial case temperature of the power device on the threshold voltage, thereby improving the accuracy of the threshold voltage degradation analysis.

[0011] In one embodiment, the process of obtaining the threshold voltage drift in step S1 includes:

[0012] When performing N short-circuit tests on n power devices with inconsistent initial case temperatures, the threshold voltage of the power device is collected once every N / m short-circuit tests to obtain m threshold voltages for each of the n power devices, where n, m and N are all positive integers and m is a preset value.

[0013] Based on the difference between each threshold voltage and the initial threshold voltage of the power device, the threshold voltage drift corresponding to each threshold voltage is determined.

[0014] In this application, by conducting N short-circuit tests on n power devices with inconsistent initial case temperatures, the threshold voltage of each power device is collected once every N / m short-circuit tests, resulting in m threshold voltages for each of the n power devices. Based on the difference between each threshold voltage and the initial threshold voltage of the power device, the threshold voltage drift corresponding to each threshold voltage is determined. In this way, based on the threshold voltage drift and the number of short circuits for each power device in each short-circuit test, the first specific value of the intermediate parameters of the threshold voltage degradation model at each initial case temperature can be obtained.

[0015] In one embodiment, the calculation process for the specific value of the model parameter K includes:

[0016] Calculate the product of each initial shell temperature with the Boltzmann constant, and the model parameters. and The difference in model parameters between them;

[0017] When the difference between any of the product results and the model parameter difference is greater than a second preset difference, the specific value of the model parameter K is determined based on the specific value of the intermediate parameter k1 at the highest initial shell temperature.

[0018] In this application, the product of each initial shell temperature and the Boltzmann constant, as well as the model parameters, are calculated. and The difference between the model parameters is used to determine the specific value of the model parameter K based on the specific value of the intermediate parameter k1 at the highest initial shell temperature, when the difference between any product result and the difference between the model parameters is greater than the second preset difference.

[0019] In one embodiment, the calculation process of the model parameter difference includes:

[0020] The expression for the intermediate parameter k1 By sequentially performing the reciprocal, rearrangement, and logarithm operations, we obtain the first fitted expression for the intermediate parameter k1 with respect to the initial shell temperature T. ;

[0021] Using the least squares method about The fitted straight line is used to obtain the slope in the first fitted expression. The specific value, and based on the slope The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. and Difference in model parameters between .

[0022] In this application, the least squares method is used. about The fitted straight line can be used to obtain the slope in the first fitted expression. The specific value, thus based on the slope The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. and Difference in model parameters between .

[0023] In one embodiment, the model parameters A1, A 2、 and The calculation process for the specific value includes:

[0024] Based on the expressions for intermediate parameters k1 and k2, a second fitted expression for k1k2 with respect to the initial shell temperature T is obtained. ;

[0025] Using the least squares method about The fitted straight line is used to obtain the intercept in the second fitted expression. Specific values ​​and slopes The specific value;

[0026] Based on the intercept From the specific values ​​of the model parameter K and the specific values ​​of the model parameter K, we obtain the specific values ​​of the model parameter A1, based on the slope. The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. The specific value;

[0027] Based on the specific values ​​of model parameter A1, the specific values ​​of model parameter A2 are obtained. The specific values ​​are used to obtain the model parameters. The specific value.

[0028] In this application, the least squares method is used. about The fitted straight line yields the intercept in the second fitted expression. Specific values ​​and slopes The specific value, so that it can be based on the intercept. From the specific values ​​of the model parameter K and the specific values ​​of the model parameter K, we obtain the specific values ​​of the model parameter A1, based on the slope. The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. The specific value of model parameter A2 is obtained based on the specific value of model parameter A1. The specific values ​​are used to obtain the model parameters. The specific value.

[0029] In one embodiment, the method further includes:

[0030] Based on the second specific value, the number of short circuits corresponding to each short circuit test, and the initial case temperature corresponding to each short circuit test, the drift amount is predicted by the threshold voltage degradation model to obtain the predicted threshold voltage drift amount corresponding to each short circuit test.

[0031] Based on the predicted threshold voltage drift and the threshold voltage drift corresponding to each of the short-circuit tests, the model loss of the threshold voltage degradation model is determined;

[0032] The model parameters of the threshold voltage degradation model are optimized based on the model loss to obtain an optimized threshold voltage degradation model, which is then used to output the threshold voltage drift of the device under analysis.

[0033] In this application, the threshold voltage degradation model is used to predict the drift amount based on a second specific value, the number of short circuits corresponding to each short circuit test, and the initial case temperature corresponding to each short circuit test. The predicted threshold voltage drift amount for each short circuit test is obtained. Based on the predicted threshold voltage drift amount and the threshold voltage drift amount for each short circuit test, the model loss of the threshold voltage degradation model is determined. Based on the model loss, the model parameters of the threshold voltage degradation model are optimized to obtain an optimized threshold voltage degradation model. This makes the output threshold voltage drift amount of the analyzed device more accurate when using the optimized threshold voltage degradation model to output the threshold voltage drift amount of the analyzed device.

[0034] In one embodiment, the process of determining the threshold voltage degradation model includes:

[0035] Acquiring a dynamic trap charge model The relationship between threshold voltage drift and the surface density of trapped electrons The relationship between the electron capture rate of the gate oxide layer and the initial case temperature in power devices. The relationship between the electron release rate of the gate oxide layer and the initial case temperature and the duration t of a single short circuit SC With the number of short circuits n SC ;

[0036] Based on the aforementioned trap charge dynamic capture model, the analytical expression for trap charge dynamic capture under short-circuit conditions is obtained. ;

[0037] Substituting the analytical formula for the dynamic trapping of charges under short-circuit conditions into the relationship between the threshold voltage drift and the surface density of trapped electrons, we obtain the threshold voltage degradation model. The definition of intermediate parameter k1 and the definition of intermediate parameter k2 ;

[0038] Substituting the relationships between electron capture rate and initial shell temperature, and between electron release rate and initial shell temperature, into the definitions of intermediate parameters k1 and k2, we obtain... and ;

[0039] Where, n t It is the surface density of electrons trapped in the trap, N n It is the electron trap surface density, c n,CIt is the electron capture rate, e n,C It is the electron release rate, q is the electron charge, and C is the electron release rate. ox It is a gate oxide capacitor, t ox It is the thickness of the gate oxide layer. It is the relative permittivity of the gate oxide layer. For the electron capture cross section, m ox It is the effective electron mass in the gate oxide layer, and h is Planck's constant. Let E be the height of the Schottky barrier and E be the electric field strength in the gate oxide layer. E is the vacuum permittivity. trap It is a trap level, E c It is the lower energy level of the conduction band.

[0040] In this application, by constructing a threshold voltage degradation model based on the dynamic evolution of trapped charge, the threshold voltage degradation is correlated with electron capture rate, electron release rate, and temperature, realizing a quantitative mapping from microscopic mechanism to macroscopic performance. This significantly improves the prediction accuracy, temperature extrapolation capability, and failure mechanism identification level of power devices under complex operating conditions, providing a solid theoretical foundation for the design, evaluation, and application of high-reliability power devices.

[0041] A threshold voltage degradation analysis apparatus for power devices, the apparatus comprising:

[0042] The test data acquisition module is used to acquire the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests.

[0043] The fitting calculation module is used to perform fitting calculations using a threshold voltage degradation model based on the threshold voltage drift and the number of short circuits of each power device in each of the short circuit tests, to obtain the first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each of the initial case temperatures.

[0044] The calculation module is used to calculate a second specific value of the model parameters in the threshold voltage degradation model based on the first specific value; the threshold voltage degradation model is... , , K, A1, A2 and All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. SC n represents the duration of a single short circuit. SC For the number of short circuits, This represents the threshold voltage drift.

[0045] The degradation analysis module is used to obtain the threshold voltage drift of the device under analysis based on the second specific value, the initial case temperature of the device under analysis, and the number of short circuits, through the threshold voltage degradation model. The device under analysis and the power device have the same model, and the initial case temperature of the device under analysis is arbitrary.

[0046] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the method described above.

[0047] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method.

[0048] The aforementioned threshold voltage degradation analysis device and computer equipment for power devices, through the use of and By determining the specific values ​​of the intermediate parameters k1 and k2 in the threshold voltage degradation model, the influence of the initial case temperature T of the device under analysis on the threshold voltage can be considered. Since the threshold voltage degradation model is related to the number of short circuits of the device under analysis, this application can simultaneously consider the influence of the number of short circuits and the initial case temperature of the power device on the threshold voltage, thereby improving the accuracy of the threshold voltage degradation analysis. Attached Figure Description

[0049] Figure 1 This is a diagram illustrating the application environment of a threshold voltage degradation analysis method for power devices in one embodiment.

[0050] Figure 2 This is a flowchart illustrating a threshold voltage degradation analysis method for power devices in one embodiment;

[0051] Figure 3 This is a schematic diagram of the overall process of a threshold voltage degradation analysis method for power devices in one embodiment;

[0052] Figure 4 This is a schematic diagram illustrating the relationship between the number of short circuits and the threshold voltage drift of a SiC MOSFET in one embodiment.

[0053] Figure 5 This is a schematic diagram illustrating the relationship between the number of short circuits and the threshold voltage drift of a SiC MOSFET in another embodiment.

[0054] Figure 6 This is a structural block diagram of a threshold voltage degradation analysis device for power devices in one embodiment. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] The threshold voltage degradation analysis method for power devices provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 interacts with server 104 via a wired / wireless channel. The data storage system can store the data that server 104 needs to process. The server obtains the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short-circuit tests.

[0057] The server, based on the threshold voltage drift and number of short-circuit tests for each power device, uses a threshold voltage degradation model for fitting calculations to obtain the first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each initial case temperature. Based on these first specific values, the server calculates the second specific values ​​of the model parameters in the threshold voltage degradation model. The threshold voltage degradation model is as follows: , , Capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference and trap level shift All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. sc n represents the duration of a single short circuit. SC For the number of short circuits, The threshold voltage drift is calculated by the server based on a threshold voltage degradation model with a pre-defined second specific value, the initial case temperature of the device under analysis, and the number of short circuits. The device under analysis and the power device have the same model number, and the initial case temperature of the device under analysis is arbitrary. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, etc. Server 104 can be a single server, a server cluster consisting of multiple servers, or a cloud computing center consisting of multiple servers.

[0058] In one embodiment, such as Figure 2 As shown, a threshold voltage degradation analysis method for power devices is provided, which can be applied to... Figure 1 Taking server 104 as an example, the following steps are included:

[0059] S1. Obtain the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests.

[0060] Each power device has the same model number. These power devices are semiconductor power devices. For example, a SiCMOSFET (Silicon Carbide Metal-Oxide-Semiconductor Field-Effect Transistor) is one such power device.

[0061] Threshold voltage drift refers to the difference between the threshold voltage of a power device and its initial threshold voltage before a short-circuit test. The initial threshold voltage is a parameter determined during the manufacturing process of the power device.

[0062] Each power device underwent multiple short-circuit tests, and the initial case temperature of each power device before the short-circuit test was different. The initial case temperature refers to the temperature of the power device's casing before the short-circuit test.

[0063] The short-circuit count is the total number of short-circuit events experienced by the power device. In each short-circuit test, the short-circuit duration of the power device is fixed.

[0064] The threshold voltage drift and the number of short circuits can be obtained from each short circuit test, or a subset of short circuit tests can be selected from all short circuit tests to obtain these values. For example, if 20 power devices are subjected to 100 short circuit tests, the threshold voltage drift and the number of short circuits in these 100 tests can be obtained. Alternatively, 5 short circuit tests can be selected from the 100 tests, and the threshold voltage drift and the number of short circuits in these 5 tests can be obtained.

[0065] S2. Based on the threshold voltage drift and number of short circuits of each power device in each short circuit test, the threshold voltage degradation model is used for fitting calculation to obtain the first specific value of the intermediate parameters of the threshold voltage degradation model at each initial case temperature.

[0066] Since each power device has a different initial case temperature, and each power device has a corresponding threshold voltage drift and number of short circuits, the threshold voltage drift and number of short circuits of a power device also correspond to its initial case temperature. For example, if the threshold voltage drift X and the number of short circuits Y of a power device with an initial case temperature A in Y short circuit tests are obtained, then the initial case temperature A, the threshold voltage drift X, and the number of short circuits Y are corresponding or related.

[0067] Specifically, based on the threshold voltage drift and the number of short circuits for each short-circuit test, and the voltage degradation model... By using the least squares method for fitting calculations, the first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each initial case temperature can be obtained. For example, based on the threshold voltage drift X and the number of short circuits Y in Y short-circuit tests of a power device with an initial case temperature of A, and the threshold voltage degradation model, the first specific values ​​of the intermediate parameters k1 and k2 at the initial case temperature A can be obtained by using the least squares method for fitting calculations.

[0068] S3. Based on the first specific value, calculate the second specific value of the model parameters in the threshold voltage degradation model;

[0069] The threshold voltage degradation model is , , Capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference and trap level shift All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. sc n represents the duration of a single short circuit. SC For the number of short circuits, This represents the threshold voltage drift.

[0070] The capacitance factor K represents the influence of charge change per unit area caused by electron traps in the oxide layer on the threshold voltage. The capacitance factor K is related to the oxide layer thickness, dielectric constant, trap density, and unit charge.

[0071] The trapping rate coefficient A1 is the rate proportionality coefficient of electrons being trapped, which is related to the trap cross section, the Boltzmann constant of the effective electron mass, and the Planck constant.

[0072] The release rate coefficient A2 is a proportionality coefficient describing the rate at which electrons are released from the trap, and it is related to the trap cross-section, the effective mass of the electron, the Boltzmann constant, and the Planck constant.

[0073] Barrier modulation energy difference This represents the difference between the height of the Schottky barrier and the barrier reduction induced by the electric field.

[0074] Trap level offset This represents the energy difference between the bottom conduction band level and the trap level.

[0075] Except for model parameters Apart from that, all other model parameters are only related to the material properties of the power device. In addition to being related to the material properties of the power device, it is also related to the electric field of the gate oxide layer, but not to the temperature.

[0076] S4. Based on the second specific value, the initial case temperature of the device under analysis, and the number of short circuits, the threshold voltage drift of the device under analysis is obtained through the threshold voltage degradation model. The device under analysis and the power device have the same model, and the initial case temperature of the device under analysis is arbitrary.

[0077] The initial case temperature of the device under analysis refers to the case temperature of the device. The number of short circuits of the device under analysis refers to the number of short circuit events that the device has experienced.

[0078] The first specific value calculated from the threshold voltage drift and the number of short-circuit tests is only the specific value of the intermediate parameter at each initial case temperature. However, the number of initial case temperatures for power devices undergoing short-circuit tests is limited and cannot cover all possible case temperatures. Therefore, based on the first specific value, it is necessary to calculate the second specific value of the model parameter in the threshold voltage degradation model to determine the relationship between the intermediate parameter and the initial case temperature of the power device. Thus, by understanding the relationship between the intermediate parameter and the initial case temperature of the power device, the specific value of the intermediate parameter at various initial case temperatures can be determined, i.e., the value of the intermediate parameter is obtained. and By obtaining the specific values ​​of each model parameter, we can obtain the expression for the intermediate parameters with respect to the initial shell temperature T. This expression can then be used to determine the specific values ​​of the intermediate parameters at various initial shell temperatures.

[0079] Specifically, the initial case temperature, capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference of the device under analysis are considered. and trap level shift Substitute the specific value into and The specific values ​​of intermediate parameters k1 and k2 at the initial case temperature of the device under analysis are obtained. The specific values ​​of intermediate parameters k1 and k2 at the initial case temperature of the device under analysis and the number of short circuits of the device under analysis are substituted into the threshold voltage degradation model to obtain the threshold voltage drift of the device under analysis.

[0080] In the above-mentioned threshold voltage degradation analysis method for power devices, by using and By determining the specific values ​​of the intermediate parameters k1 and k2 in the threshold voltage degradation model, the influence of the initial case temperature T of the device under analysis on the threshold voltage can be considered. Since the threshold voltage degradation model is related to the number of short circuits of the device under analysis, this application can simultaneously consider the influence of the number of short circuits and the initial case temperature of the power device on the threshold voltage, thereby improving the accuracy of the threshold voltage degradation analysis.

[0081] In one embodiment, the process of obtaining the threshold voltage drift in step S1 includes:

[0082] When performing N short-circuit tests on n power devices with inconsistent initial case temperatures, the threshold voltage of the power device is collected once every N / m short-circuit tests to obtain m threshold voltages for each of the n power devices. n, m and N are all positive integers, and m is a preset value.

[0083] Based on the difference between each threshold voltage and the initial threshold voltage of the power device, the threshold voltage drift corresponding to each threshold voltage is determined.

[0084] Furthermore, if N / m is not an integer, then N / m is rounded up or down, and the rounded value is determined as the number of short-circuit tests at intervals. For example, if the short-circuit test N is 100, the preset value m is 6, the number of power devices n is 10, N / m is 16.7, and rounded up to 17, then the threshold voltage of the power devices is collected in the 17th, 34th, 51st, 68th, and 85th short-circuit tests of these 10 power devices, respectively.

[0085] The number of short-circuit tests refers to which short-circuit test was conducted to collect the threshold voltage. For example, if the threshold voltage is the threshold voltage collected in the 20th short-circuit test, then the number of short-circuit tests for that threshold voltage is 20. The short-circuit duration of each short-circuit test is fixed.

[0086] The threshold voltage drift corresponding to the threshold voltage is the difference between the current threshold voltage and the initial threshold voltage.

[0087] Since the threshold voltage is collected during the short-circuit test, the number of short circuits and the threshold voltage drift corresponding to the threshold voltage are also the number of short circuits and the threshold voltage drift in the short-circuit test.

[0088] In this embodiment, by performing N short-circuit tests on n power devices with inconsistent initial case temperatures, the threshold voltage of the power device is collected once every N / m short-circuit tests, resulting in m threshold voltages for each of the n power devices. Based on the difference between each threshold voltage and the initial threshold voltage of the power device, the threshold voltage drift corresponding to each threshold voltage is determined. In this way, based on the threshold voltage drift and the number of short circuits for each power device in each short-circuit test, the first specific value of the intermediate parameter of the threshold voltage degradation model at each initial case temperature can be obtained.

[0089] In one embodiment, the calculation process for the specific value of the model parameter K includes:

[0090] Calculate the product of each initial shell temperature with the Boltzmann constant, and the model parameters. and The difference in model parameters between them;

[0091] When the difference between any product result and the difference between the model parameters is greater than the second preset difference, the specific value of the model parameter K is determined based on the specific value of the intermediate parameter k1 at the highest initial shell temperature.

[0092] Among them, model parameters and The difference in model parameters between them can be determined by and Calculation. The second preset difference is a relatively large value. Let E be the height of the Schottky barrier and E be the electric field strength in the gate oxide layer. E is the vacuum permittivity. trap It is a trap level. It is the relative permittivity of the gate oxide layer, E c It is the bottom energy level of the conduction band, and q is the electron charge.

[0093] According to the formula for calculating the intermediate parameter k1 It can be determined that when any product result kT differs from the model parameters When the difference between them is greater than the second preset difference, exp(( ) / kT)≈1, which means the higher the initial shell temperature, the higher the exp(( The closer the ratio of K to k1 is to 1, the closer the ratio of K to k1 is to a constant. Therefore, the specific value of the model parameter K can be calculated based on the specific value of the intermediate parameter k1 at the highest initial shell temperature.

[0094] Furthermore, the product of the highest initial shell temperature and the Boltzmann constant, and the model parameters were calculated. and The model parameter K is determined to be the specific value when the difference between the product result and the model parameter difference is greater than a second preset difference. k 1,T This is the specific value of the intermediate parameter k1 at the highest initial shell temperature.

[0095] In this embodiment, the product of each initial shell temperature and the Boltzmann constant, as well as the model parameters, are calculated. and The difference between the model parameters is used to determine the specific value of the model parameter K based on the specific value of the intermediate parameter k1 at the highest initial shell temperature, when the difference between any product result and the difference between the model parameters is greater than the second preset difference.

[0096] In one embodiment, the calculation process for the model parameter difference includes:

[0097] The expression for the intermediate parameter k1 By sequentially performing the reciprocal, rearrangement, and logarithm operations, we obtain the first fitted expression for the intermediate parameter k1 with respect to the initial shell temperature T. ;

[0098] Using the least squares method about The fitted straight line yields the slope in the first fitted expression. The specific value, and based on the slope The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. and Difference in model parameters between .

[0099] Specifically, the expression for the intermediate parameter k1 Take the reciprocal of both sides, multiply both sides of the equation by the model parameter K, rearrange the terms, and finally take the natural logarithm of both sides to obtain the first fitted expression of the intermediate parameter k1 with respect to the initial shell temperature T. .

[0100] Using the least squares method about The fitted line includes: As the dependent variable, As the independent variable, the first specific value of the intermediate parameter k1 at each initial shell temperature and each initial shell temperature are substituted into the equation. and In the process, multiple coordinate points are obtained, and the least squares method is used to perform linear fitting on each coordinate point to obtain... about The fitted line can be used to determine the slope in the first fitting expression. The specific value of the intermediate parameter k1 is then substituted into the initial shell temperature of each target and the initial shell temperature of each target. and In the process, multiple coordinate points are obtained, and the target initial shell temperature is the initial shell temperature excluding the highest initial shell temperature.

[0101] Due to the slope Both the Boltzmann constant k and the model parameter difference are known. It can be done by slope The specific value is obtained by multiplying it by the Boltzmann constant k.

[0102] In this embodiment, the least squares method is used. about The fitted straight line can be used to obtain the slope in the first fitted expression. The specific value, thus based on the slope The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. and Difference in model parameters between .

[0103] In one embodiment, model parameters A1, A 2、 and The calculation process for the specific value includes:

[0104] Based on the expressions for intermediate parameters k1 and k2, a second fitted expression for k1k2 with respect to the initial shell temperature T is obtained. ;

[0105] Using the least squares method about The fitted straight line yields the intercept in the second fitted expression. Specific values ​​and slopes The specific value;

[0106] Based on intercept From the specific values ​​of the model parameter K and the specific values ​​of the model parameter K, we obtain the specific values ​​of the model parameter A1, based on the slope. The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. The specific value;

[0107] Based on the specific values ​​of model parameter A1, the specific values ​​of model parameter A2 are obtained. The specific values ​​are used to obtain the model parameters. The specific value.

[0108] The process of obtaining the second fitting expression includes: an expression based on intermediate parameters k1 and k2. , ,calculate The expression is simplified, and the logarithm of both sides of the simplified equation is taken to obtain the second fitted expression. .

[0109] Using the least squares method about The process of fitting a straight line includes: As the dependent variable, As independent variables, the first specific values ​​of intermediate parameters k1 and k2 at each initial shell temperature and each initial shell temperature are substituted into the equation. and In the process, multiple coordinate points are obtained, and the least squares method is used to perform linear fitting on each coordinate point to obtain... about The fitted line can be used to obtain the intercept in the second fitting expression. Specific values ​​and slopes The specific value.

[0110] Model parameters The specific value is the slope. The specific value is the opposite of the product of the Boltzmann constant k.

[0111] The specific value of model parameter A1 is ,intercept The specific values ​​of and model parameter K are both known.

[0112] Model parameters The specific value can be based on the difference between model parameters and the model parameters. The specific value is worthwhile.

[0113] The specific value of model parameter A2 can be based on the intercept in the first fitting expression. The specific values ​​of the intercept and model parameter A1 are obtained. The intercept in the first fitting expression... The specific value calculation process is as follows: The first fitted expression... As the dependent variable, As the independent variable, the first specific value of the intermediate parameter k1 at each initial shell temperature and each initial shell temperature are substituted into the equation. and In the process, multiple coordinate points are obtained, and the least squares method is used to perform linear fitting on each coordinate point to obtain... about The fitted line can be used to determine the intercept in the first fitting expression. The specific value.

[0114] In this embodiment, the least squares method is used. about The fitted straight line yields the intercept in the second fitted expression. Specific values ​​and slopes The specific value, so that it can be based on the intercept. From the specific values ​​of the model parameter K and the specific values ​​of the model parameter K, we obtain the specific values ​​of the model parameter A1, based on the slope. The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. The specific value of model parameter A2 is obtained based on the specific value of model parameter A1. The specific values ​​are used to obtain the model parameters. The specific value.

[0115] In one embodiment, the method further includes:

[0116] Based on the second specific value, the number of short circuits corresponding to each short circuit test, and the initial case temperature corresponding to each short circuit test, the drift amount is predicted by the threshold voltage degradation model to obtain the predicted threshold voltage drift amount for each short circuit test.

[0117] Based on the predicted threshold voltage drift and threshold voltage drift corresponding to each short-circuit test, the model loss of the threshold voltage degradation model is determined.

[0118] The model parameters of the threshold voltage degradation model are optimized based on model loss to obtain the optimized threshold voltage degradation model, which is then used to output the threshold voltage drift of the device under analysis.

[0119] In a specific application, the optimized values ​​of each model parameter are shown in Table 1.

[0120] Table 1

[0121] Specifically, the second specific value and the initial case temperature corresponding to each short-circuit test are substituted into... and In this process, the specific values ​​of the intermediate parameters at each initial shell temperature can be obtained. By using the specific values ​​of the intermediate parameters at each initial shell temperature and the number of short circuits corresponding to each short circuit test, the predicted threshold voltage drift corresponding to each short circuit test can be obtained using the threshold voltage degradation model.

[0122] The model loss can be obtained based on the predicted threshold voltage drift and the sum of squared errors between the threshold voltage drift for each short-circuit test.

[0123] In this embodiment, based on the second specific value, the number of short circuits corresponding to each short circuit test, and the initial case temperature corresponding to each short circuit test, the threshold voltage degradation model is used to predict the drift amount, thereby obtaining the predicted threshold voltage drift amount corresponding to each short circuit test. Based on the predicted threshold voltage drift amount and the threshold voltage drift amount corresponding to each short circuit test, the model loss of the threshold voltage degradation model is determined. Based on the model loss, the model parameters of the threshold voltage degradation model are optimized to obtain the optimized threshold voltage degradation model. In this way, when using the optimized threshold voltage degradation model to output the threshold voltage drift amount of the device under analysis, the output threshold voltage drift amount of the device under analysis is more accurate.

[0124] In one embodiment, the process of determining the threshold voltage degradation model includes:

[0125] Acquiring a dynamic trap charge model The relationship between threshold voltage drift and the surface density of trapped electrons The relationship between the electron capture rate of the gate oxide layer and the initial case temperature in power devices. The relationship between the electron release rate of the gate oxide layer and the initial case temperature and the duration t of a single short circuit SC With the number of short circuits n SC ;

[0126] Based on the dynamic trapping model of trapped charges, the analytical expression for dynamic trapping of trapped charges under short-circuit conditions is obtained. ;

[0127] Substituting the analytical formula for the dynamic trapping of charges under short-circuit conditions into the relationship between the threshold voltage drift and the surface density of trapped electrons, we obtain the threshold voltage degradation model. The definition of intermediate parameter k1 and the definition of intermediate parameter k2 ;

[0128] Substituting the relationships between electron capture rate and initial shell temperature, and between electron release rate and initial shell temperature, into the definitions of intermediate parameters k1 and k2, we obtain... and ;

[0129] Where, n t It is the surface density of electrons trapped in the trap, N n It is the electron trap surface density, c n,C It is the electron capture rate, e n,C It is the electron release rate, q is the electron charge, and C is the electron release rate. ox It is a gate oxide capacitor, t ox It is the thickness of the gate oxide layer. It is the relative permittivity of the gate oxide layer. For the electron capture cross section, m ox It is the effective electron mass in the gate oxide layer, and h is Planck's constant. Let E be the height of the Schottky barrier and E be the electric field strength in the gate oxide layer. E is the vacuum permittivity. trap It is a trap level, E c It is the lower energy level of the conduction band.

[0130] During the short circuit, the gate-source voltage of the power device remains constant at 0. However, the thermal excitation of electrons causes a shift in the threshold voltage. This shift essentially indicates that electrons are trapped near the interface. Therefore, the dynamic trapping model for electron charges primarily needs to consider the capture and release of electrons. Based on these considerations, the dynamic trapping model for electron charges can be derived. .

[0131] When the power device is not short-circuited, the number of electrons trapped by the gate oxide of the power device is approximately zero. Therefore, by solving for and substituting the relationship between the total short-circuit time and the number of short-circuit cycles, we can obtain the analytical formula for the dynamic trapping of trapped charges under short-circuit conditions. .

[0132] Under short-circuit conditions, the current in the gate oxide layer of a power device is dominated by the thermally excited current. Therefore, the relationship between the electron capture rate of the gate oxide layer and the initial case temperature in a power device can be obtained. The relationship between the electron release rate of the gate oxide layer and the initial case temperature. .

[0133] The model parameter K is defined as follows: The model parameter A1 is defined as follows: The model parameter A2 is defined as follows: Model parameters The definition is Model parameters The definition is .

[0134] In this embodiment, by constructing a threshold voltage degradation model based on the dynamic evolution of trapped charges, the threshold voltage degradation is correlated with electron capture rate, electron release rate, and temperature, realizing a quantitative mapping from microscopic mechanism to macroscopic performance. This significantly improves the prediction accuracy, temperature extrapolation capability, and failure mechanism identification level of power devices under complex operating conditions, providing a solid theoretical foundation for the design, evaluation, and application of highly reliable power devices.

[0135] This application also provides an application scenario in which the above-described threshold voltage degradation analysis method for power devices is applied. Specifically, the application of the threshold voltage degradation analysis method for power devices in this scenario is as follows:

[0136] Step 1: Obtain experimental data. Select n identical SiC MOSFETs, ensuring that the SiC MOSFETs do not suffer destructive damage (e.g., by controlling the short-circuit time t). SC Short-circuit energy E SC To ensure that the SiC MOSFET does not suffer destructive damage, n different initial case temperatures T = (T1, T2, ... T) were applied. n A series of N repeated short-circuit experiments were conducted on each SiC MOSFET, with the same number of short-circuit tests N / m at intervals. The threshold voltage was measured periodically to obtain a dataset of threshold voltage drift measurements. By controlling the short-circuit duration and short-circuit energy, it can be ensured that the SiC MOSFET does not suffer destructive damage. (D) D represents the threshold voltage drift of a SiC MOSFET with an initial case temperature of 1 during the M×(N / m)th short-circuit test. The threshold voltage drift of a SiC MOSFET with an initial case temperature n during 1×(N / m) short-circuit tests.

[0137] Step 2: Extract intermediate parameters k1 and k2. Each threshold voltage drift corresponds to one short-circuit experiment number, thus the relationship between the threshold voltage drift and the number of short-circuit experiments can be obtained. Based on the threshold voltage degradation model, the intermediate parameters are obtained by least squares fitting for each SiC MOSFET in Step 1 at each initial case temperature T. i The specific value k1(T) i ) and k2(T i (i=1,2…n).

[0138] Step 3: Initial parameter estimation. When any initial shell temperature T and the product of this temperature and the Boltzmann constant, kT, are much greater than... When K approaches Take the k1(T) obtained in step 2. i Substitute the k1 value at the highest temperature in the equation. We obtain a preliminary estimate of K, i.e., the specific value of K. The formula for calculating the intermediate parameter k1 is then used. Take the reciprocals of both sides, multiply by K, rearrange terms, and finally take the natural logarithm of both sides to obtain the first fitted expression for the intermediate parameter k1 with respect to the initial shell temperature T. The threshold voltage offset with the highest initial case temperature is removed, and the least squares method is used to... about The fitted straight line, its intercept For a constant value The slope is And thus obtain The estimated value is the specific data of the model parameter difference. Based on the expressions for the intermediate parameters k1 and k2, a second fitted expression for k1k2 with respect to the initial shell temperature T is obtained. Using the least squares method about The fitted straight line yields the intercept in the second fitted expression. Specific values ​​and slopes The specific value of A1 and... The valuation, namely A1 and The specific values ​​of K and k are constants, thus an estimated value can be obtained, and then... and The valuation yielded A2 and Valuation.

[0139] Step 4: Parameter optimization. Using the estimated initial values ​​(K, A1, A2, ΔE1, ΔE2) from Step 3. init Starting with this, we optimize the model parameters. During optimization, we can use optimization algorithms to minimize the sum of squared residuals. Optimize, This refers to the threshold voltage drift obtained through a short-circuit test. This represents the threshold voltage drift predicted by the threshold voltage degradation model.

[0140] Step 5: Calculate the final optimized parameters (K, A1, A2, ΔE1, ΔE2) obtained in Step 4. opti Substitute into intermediate parameters and The intermediate parameters k1(T) and k2(T) are obtained. Finally, based on the initial case temperature of the SiC MOSFET to be analyzed, the specific values ​​of the intermediate parameters k1 and k2 are determined. The specific values ​​of the intermediate parameters k1 and k2 and the initial case temperature of the SiC MOSFET to be analyzed are substituted into the threshold voltage degradation model to obtain the threshold voltage drift of the SiC MOSFET to be analyzed. Figure 3 This is a flowchart illustrating the steps described above. Figure 4 and Figure 5 This diagram illustrates the relationship between the number of short circuits and the threshold voltage drift of a SiC MOSFET. The horizontal axis represents the number of short circuits, and the vertical axis represents the threshold voltage drift.

[0141] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0142] Based on the same inventive concept, this application also provides a threshold voltage degradation analysis apparatus for power devices to implement the threshold voltage degradation analysis method for power devices described above. The solution provided by this apparatus is similar to the implementation described in the above method. Therefore, the specific limitations of one or more embodiments of the threshold voltage degradation analysis apparatus for power devices provided below can be found in the limitations of the threshold voltage degradation analysis method for power devices described above, and will not be repeated here.

[0143] In one embodiment, such as Figure 6 As shown, a threshold voltage degradation analysis device for power devices is provided, comprising:

[0144] The test data acquisition module is used to acquire the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests.

[0145] The fitting calculation module is used to perform fitting calculations using a threshold voltage degradation model based on the threshold voltage drift and the number of short circuits of each power device in each of the short circuit tests, to obtain the first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each of the initial case temperatures.

[0146] The calculation module is used to calculate a second specific value of the model parameters in the threshold voltage degradation model based on the first specific value; the threshold voltage degradation model is... , , Capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference and trap level shift All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. sc n represents the duration of a single short circuit. SC For the number of short circuits, This represents the threshold voltage drift.

[0147] The degradation analysis module is used to obtain the threshold voltage drift of the device under analysis based on the second specific value, the initial case temperature of the device under analysis, and the number of short circuits, through the threshold voltage degradation model. The device under analysis and the power device have the same model, and the initial case temperature of the device under analysis is arbitrary.

[0148] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0149] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for threshold voltage degradation analysis of power devices, characterized in that, The method includes: S1. Obtain the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests. S2. Based on the threshold voltage drift and the number of short circuits of each power device in each short circuit test, a threshold voltage degradation model is used for fitting calculation to obtain the first specific value of the intermediate parameters of the threshold voltage degradation model at each initial case temperature. S3. Based on the first specific value, calculate the second specific value of the model parameters in the threshold voltage degradation model; the threshold voltage degradation model is... , , Capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference and trap level shift All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. SC n represents the duration of a single short circuit. SC For the number of short circuits, This represents the threshold voltage drift. S4. Based on the second specific value, the initial case temperature of the device to be analyzed, and the number of short circuits, the threshold voltage drift of the device to be analyzed is obtained through the threshold voltage degradation model. The device to be analyzed and the power device have the same model, and the initial case temperature of the device to be analyzed is arbitrary.

2. The method according to claim 1, characterized in that, The process of obtaining the threshold voltage drift in step S1 includes: When performing N short-circuit tests on n power devices with inconsistent initial case temperatures, the threshold voltage of the power device is collected once every N / m short-circuit tests to obtain m threshold voltages for each of the n power devices, where n, m and N are all positive integers and m is a preset value. Based on the difference between each threshold voltage and the initial threshold voltage of the power device, the threshold voltage drift corresponding to each threshold voltage is determined.

3. The method according to claim 1, characterized in that, The calculation process for the specific value of the model parameter K includes: Calculate the product of each initial shell temperature with the Boltzmann constant, and the model parameters. and The difference in model parameters between them; When the difference between any of the product results and the model parameter difference is greater than a second preset difference, the specific value of the model parameter K is determined based on the specific value of the intermediate parameter k1 at the highest initial shell temperature.

4. The method according to claim 3, characterized in that, The calculation process for the model parameter difference includes: The expression for the intermediate parameter k1 By sequentially performing the reciprocal, rearrangement, and logarithm operations, we obtain the first fitted expression for the intermediate parameter k1 with respect to the initial shell temperature T. ; Using the least squares method about The fitted straight line is used to obtain the slope in the first fitted expression. The specific value, and based on the slope The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. and Difference in model parameters between .

5. The method according to claim 3, characterized in that, The model parameters A1, A 2、 and The calculation process for the specific value includes: Based on the expressions for intermediate parameters k1 and k2, a second fitted expression for k1k2 with respect to the initial shell temperature T is obtained. ; Using the least squares method about The fitted straight line is used to obtain the intercept in the second fitted expression. Specific values ​​and slopes The specific value; Based on the intercept From the specific values ​​of the model parameter K and the specific values ​​of the model parameter K, we obtain the specific values ​​of the model parameter A1, based on the slope. The specific values ​​of the Boltzmann constant k are used to obtain the model parameters. The specific value; Based on the specific values ​​of model parameter A1, the specific values ​​of model parameter A2 are obtained. The specific values ​​are used to obtain the model parameters. The specific value.

6. The method according to claim 1, characterized in that, The method further includes: Based on the second specific value, the number of short circuits corresponding to each short circuit test, and the initial case temperature corresponding to each short circuit test, the drift amount is predicted by the threshold voltage degradation model to obtain the predicted threshold voltage drift amount corresponding to each short circuit test. Based on the predicted threshold voltage drift and the threshold voltage drift corresponding to each of the short-circuit tests, the model loss of the threshold voltage degradation model is determined; The model parameters of the threshold voltage degradation model are optimized based on the model loss to obtain an optimized threshold voltage degradation model, which is then used to output the threshold voltage drift of the device under analysis.

7. The method according to claim 1, characterized in that, The process of determining the threshold voltage degradation model includes: Acquiring a dynamic trap charge model The relationship between threshold voltage drift and the surface density of trapped electrons The relationship between the electron capture rate of the gate oxide layer and the initial case temperature in power devices. The relationship between the electron release rate of the gate oxide layer and the initial case temperature and the duration t of a single short circuit SC With the number of short circuits n SC ; Based on the aforementioned trap charge dynamic capture model, the analytical expression for trap charge dynamic capture under short-circuit conditions is obtained. ; Substituting the analytical formula for the dynamic trapping of charges under short-circuit conditions into the relationship between the threshold voltage drift and the surface density of trapped electrons, we obtain the threshold voltage degradation model. The definition of intermediate parameter k1 and the definition of intermediate parameter k2 ; Substituting the relationships between electron capture rate and initial shell temperature, and between electron release rate and initial shell temperature, into the definitions of intermediate parameters k1 and k2, we obtain... and ; Where, n t It is the surface density of electrons trapped in the trap, N n It is the electron trap surface density, c n,C It is the electron capture rate, e n,C It is the electron release rate, q is the electron charge, and C is the electron release rate. ox It is a gate oxide capacitor, t ox It is the thickness of the gate oxide layer. It is the relative permittivity of the gate oxide layer. For the electron capture cross section, m ox It is the effective electron mass in the gate oxide layer, and h is Planck's constant. Let E be the height of the Schottky barrier and E be the electric field strength in the gate oxide layer. E is the vacuum permittivity. trap It is a trap level, E c It is the lower energy level of the conduction band.

8. A threshold voltage degradation analysis device for power devices, characterized in that, The device includes: The test data acquisition module is used to acquire the threshold voltage drift and number of short circuits for each of the multiple power devices with inconsistent initial case temperatures during multiple short circuit tests. The fitting calculation module is used to perform fitting calculations using a threshold voltage degradation model based on the threshold voltage drift and the number of short circuits of each power device in each of the short circuit tests, to obtain the first specific values ​​of the intermediate parameters of the threshold voltage degradation model at each of the initial case temperatures. The calculation module is used to calculate a second specific value of the model parameters in the threshold voltage degradation model based on the first specific value; the threshold voltage degradation model is... , , Capacitance factor K, capture rate coefficient A1, release rate coefficient A2, and barrier modulation energy difference and trap level shift All are model parameters, k1 and k2 are intermediate parameters, k is the Boltzmann constant, T is the initial shell temperature, and t is the model parameter. SC n represents the duration of a single short circuit. SC For the number of short circuits, This represents the threshold voltage drift. The degradation analysis module is used to obtain the threshold voltage drift of the device under analysis based on the second specific value, the initial case temperature of the device under analysis, and the number of short circuits, through the threshold voltage degradation model. The device under analysis and the power device have the same model, and the initial case temperature of the device under analysis is arbitrary.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

Images