An evaluation method for IGBT turn-off current tail degradation

Abstract

The present invention relates to a method for evaluating the tail current degradation of an IGBT in the off state, including: collecting in real time the off-state current time series data of multiple IGBTs to be measured; performing filtering processing on the off-state current time series data of the IGBTs to be measured to obtain off-state current time series trend data; constructing a double exponential model U = a*exp(b*k) + c*exp(d*k), and calculating the model parameters of the double exponential model according to the off-state current time series trend data, where a, b, c, and d represent the model parameters of the double exponential model, and k represents the cycle period; establishing a state equation U(k) for fitting the off-state current degradation trajectory of the IGBT according to the model parameters, and evaluating the off-state current degradation of the IGBT to be measured according to the state equation U(k); wherein, the state equation U(k) is: U(k) = a*exp(b*k) + c*exp(d*k), in the formula, U represents the trend singularity value, and k represents the cycle period. The method for evaluating the tail current degradation of the IGBT in the off state according to the present invention utilizes singularity analysis to ensure the calculation accuracy while extracting the accurate trajectory of parameter degradation, and improves the speed of predicting the off-state current of the IGBT.

Description

Technical Field

[0001] The present invention belongs to the technical field of IGBT devices, and particularly relates to a method for evaluating the tail current degradation of IGBT in the off state. Background Art

[0002] IGBT devices are usually applied in high-power switching states. Their application characteristics involve multiple disciplines such as thermotics, electricity, mechanics, and materials. The research on their reliability and the optimization design are of great significance to the development of the power module failure theory and its industrialization. In-depth research on the failure mechanism of IGBT power modules, comprehensive analysis of the influence of factors such as chip characteristics, packaging materials, processes, and structures on module reliability, and the proposal of reliability protection and evaluation technologies have significant engineering application value.

[0003] IGBT usually operates under high-temperature, high-voltage, and high-current environmental conditions. Due to the large thermal stress and electrical stress it bears, the device performance will gradually degrade until it finally fails. Therefore, the common problems of IGBT failure mainly include overcurrent (such as phase-to-phase short circuit, PWM control failure short circuit) and overheat (such as hill-start, 0-100 km / h acceleration, energy recovery braking overvoltage, high-current turn-off overvoltage). Mitsubishi, Toyota, Infineon, etc. abroad have added temperature and current sensors to their respective developed IGBT modules to set a protection limit range for the safe area (overcurrent and overheat) of IGBT operating conditions.

[0004] Although domestic and foreign IGBT research and development units usually adopt the design of the safety protection domain of temperature and electrical stress, there are still frequent failures of IGBT devices, resulting in failures of power electronic equipment or even the entire machine. Therefore, an effective method for real-time detecting and evaluating the working performance of IGBT is needed. A large amount of on-site data shows that with the improvement of semiconductor technology and power electronic system optimization design technology, the current degradation of IGBT performance mainly manifests as the tail current degradation in the off state, that is, when in the off state, the current of IGBT should quickly drop to a very small value, but in reality, it needs to go through a slow state to enter the off state. This leads to a situation where IGBT will bear the impact of high voltage and high current in a certain interval. With the continuous progress of this switching state, the existence of this high-power interval provides a physical driving force for the continuous degradation of the device in the off state.

[0005] Currently, the methods for detecting and evaluating the tail current characteristics of IGBT in the off state have poor accuracy and cannot accurately evaluate the degradation performance of IGBT. Summary of the Invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a method for evaluating the tail current degradation of IGBT in the off state. The technical problems to be solved by the present invention are realized through the following technical solutions:

[0007] The present invention provides a method for evaluating the turn-off current tail degradation of an IGBT, including:

[0008] Step 1: Real-time collect the turn-off current time series data of multiple IGBTs to be measured;

[0009] Step 2: Filter the turn-off current time series data of the IGBT to be measured to obtain the turn-off current time series trend data;

[0010] Step 3: Construct a double-exponential model U = a*exp(b*k) + c*exp(d*k), and calculate the model parameters of the double-exponential model according to the turn-off current time series trend data, where a, b, c, and d represent the model parameters of the double-exponential model, and k represents the cycle period;

[0011] Step 4: Establish a state equation U(k) for fitting the turn-off current degradation trajectory of the IGBT according to the model parameters, and evaluate the turn-off current degradation of the IGBT to be measured according to the state equation U(k);

[0012] Among them, the state equation U(k) is:

[0013] U(k) = a*exp(b*k) + c*exp(d*k),

[0014] In the formula, U represents the trend singularity value, and k represents the cycle period.

[0015] In an embodiment of the present invention, the step 1 includes:

[0016] Perform multiple groups of IGBT switching current tests on the IGBT to be measured under the same port voltage but different load resistance values, and real-time collect the turn-off current time series data of multiple groups of the IGBT to be measured.

[0017] In an embodiment of the present invention, in the step 2, digital filtering technology is used to extract the trend data of the turn-off current time series data of the IGBT to be measured to obtain the turn-off current time series trend data.

[0018] In an embodiment of the present invention, in the step 3, the matrix method is used to calculate the model parameters of the double-exponential model.

[0019] In an embodiment of the present invention, the step 4 includes:

[0020] Step 4.1: Use the state equation U(k) to calculate the trend singularity value U of the IGBT to be measured;

[0021] Step 4.2: Compare the trend singularity point value U with a preset trend singularity point value threshold. If the trend singularity point value U is less than the trend singularity point value threshold, it indicates that the turn-off current characteristic of the IGBT under test degrades in this cycle period.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0023] The evaluation method for the turn-off current tail degradation of the IGBT of the present invention obtains real-time degradation information of IGBT device parameters through real-time data acquisition. At the same time, singularity analysis is used to ensure the calculation accuracy while extracting the accurate trajectory of parameter degradation, and the speed of predicting the turn-off current of the IGBT is improved.

[0024] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the drawings, is described in detail as follows. Brief Description of the Drawings

[0025] Figure 1 is a schematic diagram of an evaluation method for the turn-off current tail degradation of an IGBT provided by an embodiment of the present invention;

[0026] Figure 2 is a schematic diagram of multi-group IGBT turn-on / turn-off current timing data provided by an embodiment of the present invention;

[0027] Figure 3 is a schematic diagram of the turn-off current timing data of the IGBT for experiments provided by an embodiment of the present invention;

[0028] Figure 4 is based on the Figure 2 data obtained trend singularity point value U and trend evaluation schematic diagram;

[0029] Figure 5 is a schematic diagram of the trend singularity point value U and trend evaluation obtained based on continuous acquisition of the turn-off current provided by an embodiment of the present invention. Detailed Embodiment

[0030] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following, in conjunction with the drawings and specific embodiments, details an evaluation method for the turn-off current tail degradation of a GBT proposed according to the present invention.

[0031] The foregoing and other technical contents, features and effects of the present invention will be clearly presented in the following detailed description of the specific embodiments in conjunction with the accompanying drawings. Through the description of the specific embodiments, a more in-depth and specific understanding of the technical means and effects adopted by the present invention to achieve the predetermined purpose can be obtained. However, the accompanying drawings are only for reference and illustration, and are not used to limit the technical solution of the present invention.

[0032] Embodiment 1

[0033] IGBT belongs to a fully controlled device with self-turn-off ability, and its main current is controlled by the conductive channel in the MOS. The turn-off current of IGBT is divided into two stages: a rapid decline stage and a slow decline stage. In the first stage, which is the process of rapid current decline, when the gate voltage is removed, the channel of the MOSFET turns off and the channel current disappears. Therefore, the collector current Ic will experience a rapid decline process. In the second stage, which is the process of slow current decline, this is because there are still many electrons remaining in the n-base region (drift region), and the number depends on the current magnitude before turn-off and the thickness of the base region. To maintain electrical neutrality balance, these electrons need to recombine with holes. As the recombination process progresses, the remaining electrons become fewer and fewer, thus forming a tail current.

[0034] The collector current of IGBT is composed of the MOSFET channel current and the BJT current. When the channel turns off, the MOSFET channel current disappears, and only the BJT current remains. Therefore, the MOSFET channel current will determine the decline amplitude of the collector current in the first stage. After entering the second stage, the main component of the collector current is the recombination current driven by the remaining carriers. Therefore, the lifetime and concentration of the remaining carriers will determine the decline process of the collector current in the second stage, that is, the tail current.

[0035] When IGBT operates in a high-temperature and high-voltage environment, the carrier concentration and lifetime will gradually increase. When the internal electric field strength of the device is very large, the carriers obtain high energy and become hot carriers. When colliding with lattice atoms and ionized impurity particles, a lot of electron-hole pairs are generated, and more electron-hole pairs are generated under the positive feedback effect, which increases the risk of latch-up effect. Secondly, a part of the hot electrons with relatively large energy will generate nearby interface states when passing through the Si-SiO 2 interface. The generation of interface states may be due to the rupture of silicon-hydrogen bonds, resulting in suspended silicon bonds. The trap charges in the interface states will cause the drift of the threshold voltage, additional surface scattering, and the decrease of mobility. The hot electron charging effect is a continuous process. With the accumulation of time, the interface state effect will gradually increase, resulting in the degradation of device performance, slower switching speed, and gradually shortened remaining service life.

[0036] If the gate voltage is removed or made lower than the gate threshold voltage to discharge the gate parasitic capacitance, the inversion layer in the P-base region under the gate cannot be maintained for a long time. The disappearance of the conductive channel cuts off the source of electrons entering the N region, and no holes are injected into the N-drift region, starting the turn-off process. If a relatively small external resistance is used for gate turn-off, the gate bias voltage is reduced, resulting in a sudden drop of the device channel current to zero. However, the hole current does not stop suddenly, and the collector current continues to flow because a large amount of charge is stored in the N-drift region. The electrons withdrawn from the depletion layer diffuse towards the junction, which is forward-biased at this time. These electrons cause the holes in the substrate P+ to continue flowing into the drift region to recombine with the withdrawn electrons. As the recombination progresses, the collector current gradually decays, resulting in a prolonged turn-off current tail, that is, there is a tail current.

[0037] This embodiment proposes an evaluation method for the tail current degradation of the IGBT off-state. By measuring the real-time current data near the off-state inflection point of the IGBT with high speed and high precision, and then using the particle swarm algorithm to extract the change trend of the trend singularity point data, a quantitative evaluation of the off-state current degradation characteristics of the IGBT is achieved. Please refer to Figure 1 , Figure 1 is a schematic diagram of an evaluation method for the tail current degradation of the IGBT off-state provided by an embodiment of the present invention. As shown in the figure, the method of this embodiment includes:

[0038] Step 1: Real-time collect the off-state current time series data of multiple IGBTs to be measured;

[0039] Specifically, for the IGBTs to be measured, under the same port voltage but different load resistance values, multiple groups of IGBT switching current tests are carried out, and the off-state current time series data of multiple groups of the IGBTs to be measured are collected in real time. As Figure 2 shown, Figure 2 is a schematic diagram of the on / off-state current time series data of multiple groups of IGBTs provided by an embodiment of the present invention.

[0040] Step 2: Perform filtering processing on the off-state current time series data of the IGBT to be measured to obtain the off-state current time series trend data;

[0041] Optionally, in Step 2, digital filtering technology is used to extract the trend data of the off-state current time series data of the IGBT to be measured to obtain the off-state current time series trend data.

[0042] Since the amount of IGBT off-state current data collected in real time at high speed is large, which is not convenient for rapid calculation and device performance evaluation. Therefore, in this embodiment, digital filtering technology is adopted to extract the trend data of the off-state current time series data, obtaining the off-state current time series trend data. While retaining the time series trend, the data volume is reduced, making it easier for subsequent calculations.

[0043] Step 3: Construct a double-exponential model \(U = a\times\exp(b\times k)+c\times\exp(d\times k)\). Calculate the model parameters of the double-exponential model based on the off-state current timing trend data, where \(a\), \(b\), \(c\), and \(d\) represent the model parameters of the double-exponential model, and \(k\) represents the cycle period;

[0044] Optionally, use the matrix method to calculate the model parameters of the double-exponential model. Specifically, in this embodiment, \(a\) and \(c\) represent the model parameters obtained by curve fitting, and \(b\) and \(d\) represent the off-state current values measured twice adjacent within one cycle period (i.e., one switching cycle). For rapid iteration, the matrix method is used to calculate the model state of the trend singularity data, and this model state is expressed as \(U(k)=[a(k), b(k), c(k), d(k)]\). During the iteration process, the model parameters \(a\), \(b\), \(c\), and \(d\) are as shown in Equation (1), where \(w\) is the error between two adjacent iterative calculations. When \(w\) meets the preset iterative error or the iteration reaches the maximum number of times, the iteration is completed to obtain the model parameters of the double-exponential model.

[0045]

[0046] Step 4: Based on the model parameters, establish a state equation \(U(k)\) for fitting the off-state current degradation trajectory of the IGBT, and evaluate the off-state current degradation of the IGBT to be measured according to the state equation \(U(k)\);

[0047] Specifically, it includes:

[0048] Step 4.1: Use the state equation \(U(k)\) to calculate the trend singularity value \(U\) of the IGBT to be measured;

[0049] Step 4.2: Compare the trend singularity value \(U\) with the preset trend singularity value threshold. If the trend singularity value \(U\) is less than the trend singularity value threshold, it indicates that the off-state current characteristics of the IGBT to be measured degrade in this cycle period.

[0050] In this embodiment, the state equation \(U(k)\) is:

[0051] \(U(k)=a\times\exp(b\times k)+c\times\exp(d\times k)\) (2),

[0052] In the formula, \(U\) represents the trend singularity value, and \(k\) represents the cycle period.

[0053] It should be noted that in the actual evaluation process, the particle swarm algorithm can be used to update and optimize the model parameters of the double-exponential model. Using the updated and optimized model parameters, an updated and optimized state equation U(k) is obtained. Using this state equation U(k), a trend graph of the trend singularity value U of the IGBT to be measured is obtained, and the off-state current degradation of the IGBT to be measured is evaluated according to the trend of the trend singularity value U, so as to achieve a more accurate quantitative evaluation of the off-state current degradation characteristics of the IGBT. Specifically, the ion swarm algorithm is an existing optimization algorithm and will not be elaborated here.

[0054] The evaluation method for the off-state current tailing degradation of the IGBT in this embodiment obtains real-time degradation information of the IGBT device parameters through actual speed data acquisition. At the same time, singularity analysis ensures the accuracy of the calculation while extracting the accurate trajectory of parameter degradation, and improves the speed of predicting the off-state current of the IGBT.

[0055] Embodiment 2

[0056] Furthermore, this embodiment illustrates the effect of the evaluation method for the off-state current tailing degradation of the IGBT in Embodiment 1 through experiments. Please refer to Figure 3 , Figure 3 is a schematic diagram of the off-state current time-series data of the IGBT for experiments provided by the embodiments of the present invention. Using this off-state current time-series data, the model parameters of the double-exponential model are fitted as shown in Table 1. Using these model parameters, the state equation U(k) is obtained, and the off-state current degradation of the IGBT is evaluated.

[0057] Table 1. Fitted model parameters

[0058] a b c d -15.09 -0.01046 -35.35 0.009936

[0059] Please refer to Figure 4 and Figure 5 , Figure 4 is a schematic diagram of the trend singularity value U and trend evaluation obtained based on the Figure 2 data provided by the embodiments of the present invention; Figure 5 is a schematic diagram of the trend singularity value U and trend evaluation obtained based on continuous acquisition of the off-state current. As shown in the figure, Figure 4 and Figure 5 are respectively the evaluation results of the trend singularity value U as the off-state current detection points of the IGBT are advanced (at 400 switching cycle periods and 600 switching cycle periods). The wavy line represents the actual measurement result, and the smooth curve represents the evaluation result obtained using the evaluation method of this embodiment.

[0060] As can be seen from the figure, the difference between the two curves of the actual measurement result and the evaluation result is relatively small for the power semiconductor device applied under complex tuning conditions, that is, the error between the degradation time of the off-state current characteristic of the IGBT in the evaluation result based on the data before the prediction starting point and the degradation time of the off-state current characteristic of the actually measured IGBT is relatively low. It shows that the evaluation method for the off-state current tail degradation of the IGBT in this embodiment can quantitatively evaluate the off-state current degradation trend of the IGBT more accurately, and the relative error value between the evaluation result and the actual measurement result is shown in Table 2.

[0061] Table 2. Relative Error of Evaluation Results

[0062] Switching state period Actual failure time (h) The present invention evaluates the actual failure time (h) Relative error 400 684 695 1.58％ 600 698 695 0.43％

[0063] It should be noted that in this article, the terms "include", "comprise" or any other variant are intended to cover non-exclusive inclusion, so that an article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the article or device including the said element.

[0064] The above content is a further detailed description of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. An evaluation method for IGBT off-state current tail degradation, characterized in that, it includes: Step 1: Real-time collect the off-state current timing data of multiple IGBTs to be measured; including: performing multiple groups of IGBT switching current tests on the IGBTs to be measured with the same port voltage but different load resistance values, and real-time collecting the off-state current timing data of multiple groups of the IGBTs to be measured; the off-state current is the collector current, including the MOSFET channel current and the tail current; Step 2: Filter the off-state current timing data of the IGBTs to be measured to obtain off-state current timing trend data; Step 3: Construct a double-exponential model U = a*exp(b*k) + c*exp(d*k), and calculate the model parameters of the double-exponential model according to the off-state current timing trend data, where a, b, c, and d represent the model parameters of the double-exponential model, and k represents the cycle period; Among them, a and c represent the model parameters obtained by curve fitting, b and d represent the off-state current values measured twice adjacent within a cycle period, and the model parameters of the double-exponential model are calculated using the matrix method. During the iteration process, the model parameters a, b, c, and d are as follows: Among them, w is the error between two adjacent iterative calculations; Step 4: According to the model parameters, establish a state equation U(k) for fitting the off-state current degradation trajectory of the IGBT, and evaluate the off-state current degradation of the IGBT to be measured according to the state equation U(k); Among them, the state equation U(k) is: U(k) = a*exp(b*k) + c*exp(d*k), in the formula, U represents the value of the trend singularity point, and k represents the cycle period; The said Step 4 includes: Step 4.1: Using the state equation U(k), calculate the value of the trend singularity point U of the IGBT to be measured; Step 4.2: Compare the value of the trend singularity point U with a preset trend singularity point value threshold. If the value of the trend singularity point U is less than the trend singularity point value threshold, it indicates that the off-state current characteristics of the IGBT to be measured degrade under this cycle period.

2. The evaluation method for IGBT off-state current tail degradation according to claim 1, characterized in that, in the said Step 2, use digital filtering technology to extract the trend data of the off-state current timing data of the IGBTs to be measured to obtain the off-state current timing trend data.

Images