B10 life prediction method of IGBT under multi-failure mode physical coupling

Abstract

The application discloses a B10 life prediction method of IGBT under the physical coupling of multiple failure modes, and belongs to the technical field of IGBT life prediction, wherein the method comprises the following steps: performing a power cycle experiment of IGBT under fixed junction temperature fluctuation; fitting a bonding wire and solder layer life prediction model; measuring and fitting an IGBT on-state voltage drop and junction shell thermal resistance degradation curve; establishing a degradation constraint equation of IGBT under the physical coupling of multiple failure modes; and iteratively calculating the B10 life of IGBT under the physical coupling of multiple failure modes. The application can realize the visualization of the IGBT failure probability, is favorable for the development of IGBT reliability management, and greatly improves the reliability of the system where the IGBT is located.

Description

Technical Field

[0001] This invention belongs to the field of IGBT lifetime prediction technology, and more specifically, relates to a method for predicting the B10 lifetime of IGBTs under the physical coupling effect of multiple failure modes. Background Technology

[0002] Insulated-Gate Bipolar Transistors (IGBTs), as important power semiconductor switches in the electrical industry, are widely used in many fields, including rail transportation, smart grids, aerospace, electric vehicles, and new energy equipment, especially in clean energy and flexible DC power transmission. Because IGBTs are voltage-controlled power switches, they often withstand hundreds or even thousands of volts during operation. This high-voltage, high-power operating environment not only places high demands on the insulation performance of the IGBT module package, but also often leads to significant losses during operation. These losses are transferred outward as heat, increasing the temperature of the IGBT chip and reducing the reliability of the IGBT module.

[0003] Lifetime prediction of IGBT modules is an important aspect of reliability research. However, traditional lifetime prediction methods have the following shortcomings: (1) they do not consider the influence of random probability distribution on lifetime; (2) they only perform lifetime prediction for a specific failure model of the IGBT module. Material aging and failure are caused by cracks, and the generation and propagation of cracks are random events. Therefore, IGBT failure is also a random event, following a certain probability distribution. The average lifetime of the IGBT can be calculated through the probability distribution. However, the coupling relationship between the bonding wires and the solder layer is very complex. Currently, further research is needed on lifetime prediction of IGBT modules under this failure mode coupling effect. Summary of the Invention

[0004] In view of the above-mentioned defects or improvement needs of the existing technology, this invention proposes a B10 lifetime prediction method for IGBTs under the physical coupling effect of multiple failure modes.

[0005] To achieve the above objectives, this invention provides a method for predicting the B10 lifetime of IGBTs under the physical coupling effect of multiple failure modes, comprising:

[0006] Power cycling experiments with fixed junction temperature fluctuations were conducted on IGBTs.

[0007] A lifetime prediction model for bonding wires and solder layers was fitted based on data obtained from power cycling experiments.

[0008] Measure and fit the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curve;

[0009] Based on the lifetime prediction model of bonding leads and solder layer and the degradation curves of IGBT on-state voltage drop and junction-shell thermal resistance, a degradation constraint equation for IGBT under the physical coupling effect of multiple failure modes is established.

[0010] The degradation constraint equation is iteratively calculated to determine the B10 lifetime of the IGBT under the physical coupling of multiple failure modes.

[0011] In some alternative implementations, the power cycling test with fixed junction temperature fluctuations for the IGBT includes:

[0012] An IGBT power cycling test platform was built. During the heating phase, the IGBT module generates power loss by flowing current and using the on-state voltage drop to achieve temperature rise. During the cooling phase, water cooling is used to accelerate the cooling of the IGBT.

[0013] Set upper and lower limits for the chip junction temperature and conduct power cycling experiments with fixed junction temperature fluctuations.

[0014] In some optional implementations, the step of fitting a bonding wire and solder layer lifetime prediction model based on data obtained from power cycling experiments includes:

[0015] Multiple power cycling tests with different temperature loads were conducted to obtain failure data;

[0016] Based on failure data, we fitted lifetime prediction models for both the bonding wires and the solder layer.

[0017] In some alternative implementations, by Lifetime prediction models for the bonding wires and solder layers were fitted, where N f-bw Predicted lifetime of bonding wire, N f-sl c represents the predicted lifetime of the solder layer. bw c sl d bw and d sl T represents the fitting parameters. m These represent the chip junction temperature.

[0018] In some alternative implementations, the measurement and fitting of the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curve includes:

[0019] Set the sampling period and measure the IGBT on-state voltage drop based on the small current method;

[0020] Measure the temperature of the IGBT module chip and the outer side of the substrate, and calculate the junction-to-case thermal resistance of the IGBT module based on the power loss.

[0021] Record the on-state voltage drop and junction thermal resistance for each sampling period until failure. Plot the degradation data of on-state voltage drop and thermal resistance as a degradation curve and fit it.

[0022] In some alternative implementations, by Obtain the junction-to-case thermal resistance R of the IGBT module th , among which, U ces For the on-state voltage drop, T j For the chip junction temperature, T c For the substrate case temperature, I c This is the collector current;

[0023] Depend on Fit the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curves, where U ces-0 and R th-0 U ces and R th The initial value of a bw a sl b bw b sl Let D be the parameter to be fitted. bw With D sl These represent the aging degree of the bonding wires and the solder layer, respectively. N represents the number of IGBT power cycles.

[0024] In some optional implementations, establishing the degradation constraint equations for the IGBT under the physical coupling of multiple failure modes includes:

[0025] Based on the lifetime prediction model of the bonding wires and solder layer, and the IGBT on-state voltage drop and junction-shell thermal resistance degradation curves, the relevant parameters of the constraint equation are obtained. These parameters include: a represents the non-exponential term of the Less lifetime prediction model; b represents the negative exponent of the exponential term of the Less lifetime prediction model; c represents the initial absolute temperature; and d represents U... ces The constant coefficient of the exponential term in the curve of change, e, represents U. ces The exponent of the exponent term in the change curve, f, represents D. bw The constant coefficients and conversion coefficients during the conversion, g represents D bw Exponential conversion factor and h representing R during conversion th The exponent of the exponential term in the change curve;

[0026] The parameters related to the constraint equations, including the initial value of the chip junction temperature and the ambient temperature, are measured, and the degradation constraint equations are established.

[0027] In some alternative implementations, by Establish the degradation constraint equation, where r a-bw This indicates the aging rate of the bonding wire.

[0028] In some optional implementations, the iterative calculation of the degradation constraint equation for the B10 lifetime of the IGBT under the physical coupling of multiple failure modes includes:

[0029] The degradation constraint equation is iteratively calculated to obtain the bonding wire and solder layer D. bw With D sl Change curve;

[0030] Based on bonding wires and solder layer D bw With D sl The variation curve is used to calculate the variation curve of the failure probability of the bonding wire and solder layer.

[0031] Calculate the overall failure probability of the IGBT module based on the failure probabilities of the bonding wires and solder layer;

[0032] The point where the overall failure probability is 10% corresponds to the number of cycles of the IGBT's B10 lifetime.

[0033] In some alternative implementations, by Calculate the failure probability variation curves of the bonding wires and solder layer, where m bw and m sl F represents the shape parameter in the Weibull distribution of the bonding wire and solder layer. bw (D bw F represents the cumulative failure rate of the bonding wire. sl (D sl This indicates the cumulative failure rate of the solder layer;

[0034] By F IGBT (N)=1-[1-F bw (D bw )][1-F sl (D sl )] Calculate the overall failure probability of the IGBT module, where F IGBT (N) represents the overall cumulative failure rate of the IGBT module.

[0035] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0036] This invention provides a method for predicting the B10 lifetime of IGBTs under the physical coupling effect of multiple failure modes. By monitoring the degradation degree of the bonding leads and solder layer, the failure probability of both is assessed, and the cumulative failure rate curve of the entire IGBT module is calculated. The number of cycles at which the cumulative failure rate reaches 10% is taken as the B10 lifetime of the IGBT module. Compared with existing technologies, this invention has advantages such as high prediction accuracy and low failure rate at the rated life of the IGBT. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of an IGBT power cycling experiment provided in an embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of temperature control provided in an embodiment of the present invention;

[0039] Figure 3 This is a schematic diagram of the B10 lifetime establishment process of an IGBT provided in an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0041] This invention provides a method for predicting the B10 lifetime of an IGBT under the physical coupling of multiple failure modes, comprising the following steps:

[0042] (1) Power cycling experiments with fixed junction temperature fluctuations were conducted on IGBTs;

[0043] (2) Fit a lifetime prediction model for bonding wires and solder layers based on data obtained from power cycling experiments;

[0044] (3) Measure and fit the IGBT on-state voltage drop and junction-shell thermal resistance degradation curves;

[0045] (4) Based on the bonding wire and solder layer lifetime prediction model and the IGBT on-state voltage drop and junction-shell thermal resistance degradation curve, the degradation constraint equation of IGBT under the physical coupling of multiple failure modes is established.

[0046] (5) Iteratively calculate the B10 lifetime of IGBT under the physical coupling of multiple failure modes by performing degradation constraint equations.

[0047] Furthermore, in step (1), the specific method for conducting a power cycling experiment with fixed junction temperature fluctuations on the IGBT is as follows:

[0048] (1-1) An IGBT power cycling experimental platform was built. During the heating phase, a large current was flowed through the IGBT module, and sufficient power loss was generated by the on-state voltage drop to achieve the temperature rise. During the cooling phase, water cooling was used to accelerate the cooling of the IGBT. The principle of the experimental platform is as follows: Figure 1 As shown;

[0049] (1-2) Set the upper and lower limits of the chip junction temperature. Control the IGBT chip junction temperature through a temperature controller and a relay. When the chip junction temperature reaches the upper limit T... jmax When the temperature controller disconnects the relay, the IGBT turns off, and heating stops, while the water cooler begins cooling the IGBT. When the chip junction temperature reaches the lower limit T... jmin When the temperature controller reconnects the relay, the water cooler stops working, and the IGBT turns on, restarting heating. Figure 2 As shown;

[0050] (1-3) Conduct a power cycling experiment with fixed junction temperature fluctuations and record the on-state voltage drop U. ces Increasing the number of cycles by 5% results in increased bond wire lifetime and thermal resistance R. th Increase the number of cycles by 20% to extend the solder layer life.

[0051] Furthermore, in step (2), the specific method for fitting the lifetime prediction model of the bonding wire and solder layer is as follows:

[0052] (2-1) Conduct multiple sets of power cycling experiments with different temperature loads to obtain failure data;

[0053] (2-2) Based on failure data, fit the lifetime prediction models of the bonding wires and solder layers respectively.

[0054] The life prediction model is shown in the following formula (1).

[0055]

[0056] Where, N f-bw N f-sl These represent the predicted lifetimes of the bonding wires and the solder layer, respectively. bw c sl d bw d sl For the fitting parameters, T m This indicates the chip junction temperature.

[0057] Furthermore, in step (3), the specific method for measuring and fitting the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curve is as follows:

[0058] (3-1) Set the sampling period to T;

[0059] (3-2) Measuring IGBT on-state voltage drop based on the small current method: Place the IGBT on a dedicated test platform, with the IGBT in the on state, and allow a small current of 100mA to flow through it. Measure the collector-emitter voltage U at this time. ces , which is the value of the on-state voltage drop;

[0060] (3-3) Measure the temperatures of the IGBT module chip and the outer side of the substrate, and record them as chip junction temperature T. j and substrate case temperature T c Combined with power loss Ploss, i.e., on-state voltage drop U ces With collector current I c The product of is used to calculate the junction-to-case thermal resistance of the IGBT module, as shown in the following formula (2):

[0061]

[0062] (3-4) Record the on-state voltage drop and thermal resistance for each sampling period until failure. Plot the degradation data of on-state voltage drop and thermal resistance as a degradation curve, and fit it using the following formula (3):

[0063]

[0064] Among them, U ces-0 and R th-0 U ces and R th The initial value of a bw a sl b bw b sl Let D be the parameter to be fitted. bw With D sl The aging degree of the bonding wire and the solder layer can be calculated using the following formula (4):

[0065]

[0066] Where N is the number of IGBT power cycles.

[0067] Furthermore, in step (4), the specific method for establishing the degradation constraint equation of the IGBT under the physical coupling of multiple failure modes is as follows:

[0068] (4-1) Based on the lifetime prediction model of bonding wires and solder layer, the on-state voltage drop of IGBT and the degradation curve of junction thermal resistance, the relevant parameters of the constraint equation are obtained, as shown in Table 1. A total of 8 parameters ah are calculated.

[0069] Table 1 Summary of methods for calculating coefficients of degradation constraint equations

[0070]

[0071] (4-2) Measure the initial value of the chip junction temperature, ambient temperature and other relevant parameters of the constraint equation;

[0072] (4-3) Establish the degradation constraint equation, which is shown in the following formula (5):

[0073]

[0074] Where, r a-bw This represents the aging rate of the bonded wire, and its value is the reciprocal of the predicted lifetime, i.e., r. a-bw =1 / N f-bw .

[0075] Furthermore, in step (5), the specific method for iteratively calculating the B10 lifetime of the IGBT under the physical coupling effect of multiple failure modes is as follows:

[0076] (5-1) Iterative calculations are performed on the constraint equations to obtain the degradation curves of the bonded leads and solder layers, i.e., D. bw With D sl Change curve;

[0077] (5-2) Based on the degradation curve, the failure probability change curve of the bonding wire and solder layer is calculated. The calculation formula is shown in (6):

[0078]

[0079] Where, m bw and m sl The shape parameters of the bonding wires and the Weibull distribution in the solder layer can be obtained by looking up a table or through fatigue testing. F bw (D bw ), F sl (D sl These represent the cumulative failure rates of the bonding wires and the solder layer, respectively.

[0080] (5-3) Based on the failure probabilities of the bonding leads and solder layer, calculate the overall failure probability of the IGBT module. The calculation formula is as follows:

[0081] F IGBT (N)=1-[1-F bw (D bw )][1-F sl (D sl (7)

[0082] Among them, F IGBT (N) represents the overall cumulative failure rate of the IGBT module.

[0083] (5-4) Take the point where the cumulative failure rate is 10%, and the corresponding number of cycles is the B10 lifetime of the IGBT. That is, take F. IGBT When (N) = 0.1, N is taken as the B10 lifetime of the IGBT.

[0084] like Figure 3 As shown, the specific process of this embodiment is as follows:

[0085] Step 01: Measure the initial value U of the IGBT characteristic parameter ces-0 and R th-0 ;

[0086] Step 02: Conduct power cycling experiments on the IGBT;

[0087] Step 03: Fit the lifetime prediction model for bonding leads and solder layer;

[0088] Step 04: Obtain the constant coefficient c bw c sl d bw d sl ;

[0089] Step 05: Record and fit U ces R th Degradation curve;

[0090] Step 06: Obtain the constant coefficient a bw a sl b bw b sl ;

[0091] Step 07: Calculate the constant coefficient ag based on the initial temperature;

[0092] Step 08: Construct the aging constraint equations for the bonding wires and solder layer;

[0093] Step 09: Calculate D iteratively bw and D sl Degradation curve;

[0094] Step 10: Calculate the overall cumulative failure rate curve of the IGBT module;

[0095] Step 11: Take the number of cycles when the cumulative failure rate is 0.1 as the B10 lifetime;

[0096] Step 12: End.

[0097] This invention provides a method for predicting the B10 lifetime of an IGBT under the physical coupling effect of multiple failure modes. By monitoring the degradation degree of the bonding leads and solder layer, the failure probability of both is assessed, and the cumulative failure rate curve of the entire IGBT module is calculated. The number of cycles at which the cumulative failure rate reaches 10% is taken as the B10 lifetime of the IGBT module. Compared with existing technologies, this invention has advantages such as high prediction accuracy and low failure rate at the rated life of the IGBT.

[0098] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0099] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the B10 lifetime of an IGBT under the physical coupling effect of multiple failure modes, characterized in that, include: Power cycling experiments with fixed junction temperature fluctuations were conducted on IGBTs. A lifetime prediction model for bonded leads and solder layers was fitted based on data obtained from power cycling experiments, wherein... Lifetime prediction models for both the bonding wires and the solder layer were fitted, where... This indicates the predicted lifetime of the bonding wire. Indicates the predicted lifetime of the solder layer. , , and Represents the fitted parameters, Indicates the chip junction temperature; Measure and fit the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curve; Based on the lifetime prediction model of bonding leads and solder layer, and the degradation curves of IGBT on-state voltage drop and junction-shell thermal resistance, a degradation constraint equation for IGBT under the physical coupling of multiple failure modes is established. Among these equations, the following equations are derived: Obtain the junction-to-shell thermal resistance of the IGBT module ,in, U ces For conduction voltage drop, T j For chip junction temperature, T c For the substrate shell temperature, I c The collector current; by Fit the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curves, where, U ces-0 and R th-0 They are respectively U ces and R th initial value, a bw , a sl , b bw , b sl The parameters to be fitted are... D bw and D sl These represent the aging degree of the bonding wires and the solder layer, respectively. , N The number of IGBT power cycles, determined by... Establish the degradation constraint equation, where, Indicates the aging rate of the bonding wire. This indicates the non-exponential term in the Less life prediction model. b The negative value of the exponent in the Less life prediction model's exponent term indicates the number of exponents. c Indicates the initial absolute temperature, d express U ces The constant coefficient of the exponential term of the change curve e express U ces The index of the index term of the change curve f express D bw Conversion coefficients during conversion g express D bw Exponential conversion factor during conversion and h express R th The exponent of the exponential term in the change curve; The degradation constraint equation is iteratively calculated to determine the B10 lifetime of the IGBT under the physical coupling of multiple failure modes.

2. The method according to claim 1, characterized in that, The power cycling experiment for IGBTs with fixed junction temperature fluctuations includes: An IGBT power cycling test platform was built. During the heating phase, the IGBT module generates power loss by flowing current and using the on-state voltage drop to achieve temperature rise. During the cooling phase, water cooling is used to accelerate the cooling of the IGBT. Set upper and lower limits for the chip junction temperature and conduct power cycling experiments with fixed junction temperature fluctuations.

3. The method according to claim 1 or 2, characterized in that, The lifetime prediction model for bonded leads and solder layers based on data obtained from power cycling experiments includes: Multiple power cycling tests with different temperature loads were conducted to obtain failure data; Based on failure data, we fitted lifetime prediction models for both the bonding wires and the solder layer.

4. The method according to claim 3, characterized in that, The measurement and fitting of the IGBT on-state voltage drop versus junction-shell thermal resistance degradation curve includes: Set the sampling period and measure the IGBT on-state voltage drop based on the small current method; Measure the temperature of the IGBT module chip and the outer side of the substrate, and calculate the junction-to-case thermal resistance of the IGBT module based on the power loss. Record the on-state voltage drop and junction thermal resistance for each sampling period until failure. Plot the degradation data of on-state voltage drop and thermal resistance as a degradation curve and fit it.

5. The method according to claim 4, characterized in that, The establishment of the degradation constraint equations for IGBTs under the physical coupling of multiple failure modes includes: Based on the lifetime prediction model of bonding wires and solder layers, and the IGBT on-state voltage drop and junction-shell thermal resistance degradation curves, the relevant parameters of the constraint equation are obtained. The parameters related to the constraint equations, including the initial value of the chip junction temperature and the ambient temperature, are measured, and the degradation constraint equations are established.

6. The method according to claim 5, characterized in that, The iterative calculation of the B10 lifetime of the IGBT under the physical coupling of multiple failure modes by the degradation constraint equation includes: The degradation constraint equation is iteratively calculated to obtain the bonding wires and solder layer. D bw and D sl Change curve; Based on bonding wires and solder layer D bw and D sl The variation curve is used to calculate the variation curve of the failure probability of the bonding wire and solder layer. Calculate the overall failure probability of the IGBT module based on the failure probabilities of the bonding wires and solder layer; The point where the overall failure probability is 10% corresponds to the number of cycles of the IGBT's B10 lifetime.

7. The method according to claim 6, characterized in that, Depend on Calculate the failure probability variation curves of the bonding wires and solder layer, where... m bw and m sl The shape parameters in the Weibull distribution of the bonding leads and solder layer. This indicates the cumulative failure rate of the bonding wires. Indicates the cumulative failure rate of the solder layer; Depend on Calculate the overall failure probability of the IGBT module, where, This indicates the overall cumulative failure rate of the IGBT module.

Images