CN117852462A - Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium - Google Patents
Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium Download PDFInfo
- Publication number
- CN117852462A CN117852462A CN202311550020.1A CN202311550020A CN117852462A CN 117852462 A CN117852462 A CN 117852462A CN 202311550020 A CN202311550020 A CN 202311550020A CN 117852462 A CN117852462 A CN 117852462A
- Authority
- CN
- China
- Prior art keywords
- simulation
- interference signal
- preset
- transistor
- reliability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 362
- 238000000034 method Methods 0.000 title claims abstract description 75
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 230000000737 periodic effect Effects 0.000 claims abstract description 173
- 239000000203 mixture Substances 0.000 claims abstract description 171
- 230000015556 catabolic process Effects 0.000 claims abstract description 127
- 238000006731 degradation reaction Methods 0.000 claims abstract description 127
- 230000001133 acceleration Effects 0.000 claims abstract description 74
- 230000006870 function Effects 0.000 claims description 116
- 230000005684 electric field Effects 0.000 claims description 66
- 238000009826 distribution Methods 0.000 claims description 63
- 238000005315 distribution function Methods 0.000 claims description 44
- 238000012360 testing method Methods 0.000 claims description 44
- 230000003068 static effect Effects 0.000 claims description 31
- 238000006243 chemical reaction Methods 0.000 claims description 26
- 238000004422 calculation algorithm Methods 0.000 claims description 17
- 238000002474 experimental method Methods 0.000 claims description 13
- 239000004065 semiconductor Substances 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 10
- 229910044991 metal oxide Inorganic materials 0.000 claims description 7
- 150000004706 metal oxides Chemical class 0.000 claims description 7
- 230000010354 integration Effects 0.000 claims description 6
- 238000009792 diffusion process Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 20
- 230000000694 effects Effects 0.000 description 19
- 230000008569 process Effects 0.000 description 14
- 230000007246 mechanism Effects 0.000 description 8
- 238000007476 Maximum Likelihood Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 230000001052 transient effect Effects 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- INQLNSVYIFCUML-QZTLEVGFSA-N [[(2r,3s,4r,5r)-5-(6-aminopurin-9-yl)-3,4-dihydroxyoxolan-2-yl]methoxy-hydroxyphosphoryl] [(2r,3s,4r,5r)-5-(4-carbamoyl-1,3-thiazol-2-yl)-3,4-dihydroxyoxolan-2-yl]methyl hydrogen phosphate Chemical compound NC(=O)C1=CSC([C@H]2[C@@H]([C@H](O)[C@@H](COP(O)(=O)OP(O)(=O)OC[C@@H]3[C@H]([C@@H](O)[C@@H](O3)N3C4=NC=NC(N)=C4N=C3)O)O2)O)=N1 INQLNSVYIFCUML-QZTLEVGFSA-N 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000000630 rising effect Effects 0.000 description 3
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000011960 computer-aided design Methods 0.000 description 2
- 230000005672 electromagnetic field Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- OBNDGIHQAIXEAO-UHFFFAOYSA-N [O].[Si] Chemical compound [O].[Si] OBNDGIHQAIXEAO-UHFFFAOYSA-N 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 229910052681 coesite Inorganic materials 0.000 description 1
- 229910052906 cristobalite Inorganic materials 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000005669 field effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 239000000377 silicon dioxide Substances 0.000 description 1
- 235000012239 silicon dioxide Nutrition 0.000 description 1
- 229910052682 stishovite Inorganic materials 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 229910052905 tridymite Inorganic materials 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a reliability simulation method, a reliability simulation device, a simulation apparatus, and a nonvolatile computer-readable storage medium. The method comprises the steps of obtaining a periodic interference signal of an electromagnetic interference signal; performing Gaussian mixture simulation on the periodic interference signal of the acquired electromagnetic interference signal to acquire a Gaussian mixture interference signal time domain function; converting the periodic interference signal into a rectangular pulse signal according to a Gaussian mixture interference signal time domain function and a preset acceleration factor model, wherein the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the preset acceleration stress condition, and meets the degradation consistency condition; and taking the rectangular pulse signal as an input boundary condition, inputting the rectangular pulse signal into preset simulation software for simulation for preset times to output the degradation rate of the transistor to be simulated, wherein the degradation rate is configured to represent the reliability of the transistor to be simulated. Thus, the simulation result can accurately reflect the reliability of the transistor to be simulated.
Description
Technical Field
The present invention relates to the field of transistor simulation technology, and more particularly, to a reliability simulation method, a reliability simulation apparatus, a simulation device, and a nonvolatile computer-readable storage medium.
Background
Currently, power transistors are widely used in circuit structures of industrial chips as semiconductor representative devices. Simulation software, such as semiconductor process simulation and device simulation software (Technology Computer Aided Design, TCAD), is used to simulate the reliability of the transistor prior to actual flow, and the transistor lifetime is estimated based on the calculation results. Whereas prior art solutions rely mainly on signal parameters of simpler pulses as transient simulation inputs in the reliability evaluation of transistors under electromagnetic fields, for example for laterally diffused metal oxide semiconductor transistors, currently periodic single transmission line pulses (Transmission Line Pulse, TLP) are often relied on as transient simulation inputs. The simulation conditions are simpler, and the common complex electromagnetic pulse environment of the industrial chip is difficult to cover, so that the reliability life expectancy value of the transistor and the actual experience value have huge differences, and the reliability of the transistor is difficult to be accurately estimated.
Disclosure of Invention
The embodiment of the application provides a reliability simulation method, a reliability simulation device, simulation equipment and a nonvolatile computer readable storage medium, which can realize equivalent conversion of complex electromagnetic interference signals to obtain rectangular pulse signals, and then realize simulation by using the rectangular pulse signals so as to improve the reliability of simulation and reduce the simulation time.
The reliability simulation method comprises the steps of obtaining a periodic interference signal of an electromagnetic interference signal, wherein the electromagnetic interference signal is formed by superposing a plurality of electromagnetic pulse signals; performing Gaussian mixture simulation on the periodic interference signal to obtain a Gaussian mixture interference signal time domain function; converting the periodic interference signal into a rectangular pulse signal according to the Gaussian mixture interference signal time domain function and a preset acceleration factor model, wherein the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the preset acceleration stress condition, and the preset acceleration factor model meets the degradation consistency condition, and the degradation consistency condition is that the degradation of the dynamic stress in the preset time period and the effective electric field static stress obtained through integration is consistent with the degradation of the generation of the dynamic stress; and taking the rectangular pulse signal as an input boundary condition, inputting the rectangular pulse signal into preset simulation software for simulation for preset times to output the degradation rate of the transistor to be simulated, wherein the degradation rate is configured to characterize the reliability of the transistor to be simulated.
In some embodiments, the converting the periodic interference signal into a rectangular pulse signal according to the gaussian mixture interference signal time domain function and a preset acceleration factor model includes: determining the preset acceleration factor model according to preset stress parameters of the transistor to be simulated; acquiring equivalent pulse voltage according to degradation consistency conditions, the preset acceleration factor model, the Gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective starting moment of the periodic interference signal and the effective ending moment of the periodic interference signal; and determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
In some embodiments, the determining the preset acceleration factor model according to the preset stress parameter of the transistor to be simulated includes: performing an accelerating electric field experiment on a plurality of test transistors at constant temperature, and obtaining the actual failure time of each test transistor by applying constant electric fields with different intensities to each test transistor, wherein the specification and the model of the plurality of test transistors are the same; taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated; acquiring preset stress parameters according to the failure time model and a group of data; and acquiring the preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
In certain embodiments, the degenerate consistency condition comprises:wherein P is the preset duration, < >>For the preset acceleration factor model, t is time, and the preset duration is according to the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signalAnd (5) engraving and determining.
In some embodiments, the acquiring the periodic interference signal of the electromagnetic interference signal includes: acquiring a signal characteristic period of the electromagnetic interference signal; and acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
In some embodiments, the performing gaussian mixture simulation on the periodic interference signal to obtain a gaussian mixture interference signal time domain function includes: acquiring a target Gaussian mixture model corresponding to the wave crest number according to the wave crest number of the periodic interference signal; and performing Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm to obtain a Gaussian mixture interference signal time domain function.
In some embodiments, the performing gaussian mixture simulation on the periodic interference signal according to the target gaussian mixture model by using a desired maximum algorithm to obtain a gaussian mixture interference signal time domain function includes: converting the voltage intensity in the time domain of the periodic interference signal into probability density in the time domain, wherein the probability density is determined according to the ratio of the voltage intensity value to the total voltage intensity value; obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the target Gaussian mixture model according to preset distribution data, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight, distribution mean value and standard deviation of each Gaussian distribution function; and generating the Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
In some embodiments, the calculating, according to preset distribution data, the optimal parameter of the gaussian mixture model probability function corresponding to the target gaussian mixture model includes: different parameter combinations are obtained, wherein the parameter combinations comprise weight, distribution mean and standard deviation of each Gaussian distribution function; calculating the distribution probability of the preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability; and determining a target parameter combination, wherein the target parameter combination is the parameter combination with the error value of the parameter combination before and after updating being smaller than a preset error value.
In some embodiments, the inputting the rectangular pulse signal as an input boundary condition into a preset simulation software for simulation for a preset number of times to output the degradation rate of the transistor to be simulated includes: determining a plurality of simulation points in the rectangular pulse signal; sequentially inputting voltages corresponding to the simulation points as gate voltages of the transistor to be simulated, and performing simulation for preset times, wherein the simulation is performed once when the input of the simulation points of the rectangular pulse signal is completed; and detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
In some embodiments, the degradation rate is characterized by a first ratio of the present drain current and the initial drain current, or the degradation rate is characterized by a second ratio of a difference between the initial drain current and the present drain current to the initial drain current.
In some embodiments, the number of simulation points when simulating with the rectangular pulse signal is smaller than the number of simulation points when simulating with the periodic disturbance signal.
In some embodiments, the simulation software performs simulation according to a preset reliability simulation model, and the reliability simulation model is determined according to the type of the transistor to be simulated.
In some embodiments, the transistor to be simulated comprises a laterally diffused metal oxide semiconductor transistor and the reliability simulation model comprises a reactive diffusion model.
In some embodiments, a plurality of the electromagnetic pulse signals are generated by operating a plurality of gas-insulated switchgear devices in parallel at the same frequency.
In some embodiments, the transistor to be emulated comprises a diode or a triode.
The reliability simulation device of the embodiment of the application comprises an acquisition module, an analog module, a conversion module and a simulation module. The acquisition module is used for acquiring a periodic interference signal of an electromagnetic interference signal, and the electromagnetic interference signal is formed by superposition of a plurality of electromagnetic pulse signals. The simulation module is used for carrying out Gaussian mixture simulation on the periodic interference signal so as to obtain a Gaussian mixture interference signal time domain function; the conversion module is used for converting the periodic interference signal into a rectangular pulse signal according to the Gaussian mixture interference signal time domain function and a preset acceleration factor model, the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the condition of preset acceleration stress, the preset acceleration factor model meets the degradation consistency condition, the degradation consistency condition is that the degradation of the dynamic stress in the preset time length, and the effective electric field static stress obtained by integration and the generation of the dynamic stress are consistent; the simulation module is used for inputting the rectangular pulse signal as an input boundary condition to preset simulation software for simulation for preset times to output the degradation rate of the transistor to be simulated, wherein the degradation rate is configured to represent the reliability of the transistor to be simulated.
In certain embodiments, the conversion module is further configured to: determining the preset acceleration factor model according to preset stress parameters of the transistor to be simulated; acquiring equivalent pulse voltage according to degradation consistency conditions, the preset acceleration factor model, the Gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective starting moment of the periodic interference signal and the effective ending moment of the periodic interference signal; and determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
In certain embodiments, the conversion module is further configured to: performing an accelerating electric field experiment on a plurality of test transistors at constant temperature, and obtaining the actual failure time of each test transistor by applying constant electric fields with different intensities to each test transistor, wherein the specification and the model of the plurality of test transistors are the same; taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated; acquiring preset stress parameters according to the failure time model and a group of data; and acquiring the preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
In certain embodiments, the degenerate consistency condition comprises:wherein P is the preset duration, < >>And t is time for the preset acceleration factor model, and the preset duration is determined according to the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
In some embodiments, the acquiring module is further configured to acquire a signal characteristic period of the electromagnetic interference signal; and acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
In some embodiments, the simulation module is further configured to obtain a target gaussian mixture model corresponding to the number of peaks of the periodic interference signal according to the number of peaks; and performing Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm to obtain a Gaussian mixture interference signal time domain function.
In some embodiments, the simulation module is further configured to convert the voltage intensity in the time domain of the periodic interference signal into a probability density in the time domain, where the probability density is determined according to a ratio of the voltage intensity value and the total voltage intensity value; obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the target Gaussian mixture model according to preset distribution data, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight, distribution mean value and standard deviation of each Gaussian distribution function; and generating the Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
In some embodiments, the simulation module is further configured to obtain different parameter combinations, where the parameter combinations include a weight, a distribution mean, and a standard deviation of each of the gaussian distribution functions; calculating the distribution probability of the preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability; and determining a target parameter combination, wherein the target parameter combination is the parameter combination with the error value of the parameter combination before and after updating being smaller than a preset error value.
In some embodiments, the simulation module is further configured to determine a plurality of simulation points in the rectangular pulse signal; sequentially inputting voltages corresponding to the simulation points as gate voltages of the transistor to be simulated, and performing simulation for preset times, wherein the simulation is performed once when the input of the simulation points of the rectangular pulse signal is completed; and detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
In some embodiments, the degradation rate is characterized by a first ratio of the present drain current and the initial drain current, or the degradation rate is characterized by a second ratio of a difference between the initial drain current and the present drain current to the initial drain current.
In some embodiments, the number of simulation points when simulating with the rectangular pulse signal is smaller than the number of simulation points when simulating with the periodic disturbance signal.
In some embodiments, the simulation software performs simulation according to a preset reliability simulation model, and the reliability simulation model is determined according to the type of the transistor to be simulated.
In some embodiments, the transistor to be simulated comprises a laterally diffused metal oxide semiconductor transistor and the reliability simulation model comprises a reactive diffusion model.
In some embodiments, a plurality of the electromagnetic pulse signals are generated by operating a plurality of gas-insulated switchgear devices in parallel at the same frequency.
In some embodiments, the transistor to be emulated comprises a diode or a triode.
The simulation device of the embodiment of the present application includes a memory and a processor, where the memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to execute the reliability simulation method of any one of the embodiments.
The non-transitory computer readable storage medium of the present embodiment includes a computer program that, when executed by a processor, causes the processor to execute the reliability simulation method of any of the above embodiments.
According to the reliability simulation method, the reliability simulation device, the simulation equipment and the nonvolatile computer readable storage medium, firstly, electromagnetic interference signals formed by superposition of a plurality of electromagnetic pulse signals are obtained, and then the electromagnetic interference signals are preprocessed to obtain periodic interference signals of the electromagnetic interference signals. And then, carrying out Gaussian mixture simulation on the periodic interference signal, and obtaining a Gaussian mixture interference signal time domain function corresponding to the periodic interference signal after simulation so as to describe the complex electromagnetic interference signal, so that the follow-up reliability simulation can be carried out based on the Gaussian mixture interference signal time domain function, and the simulation effect is ensured to be closer to the actual degradation effect. Then, an acceleration factor model meeting the degradation consistency condition can be preset, and the periodic interference signal is converted according to the preset acceleration factor model to obtain a rectangular pulse signal. At this time, the stress generated by the periodic disturbance signal may be regarded as dynamic stress, and the stress generated by the converted rectangular pulse signal may be regarded as effective electric field static stress. The rectangular pulse signal retains the rising and falling characteristics of the signal, the degradation generated by the static stress and the dynamic stress of the effective electric field is consistent, and the characteristic of the electromagnetic interference signal can be represented by combining the periodic interference signal, so that the degradation condition of the transistor is consistent when the rectangular pulse signal is input to simulation software and the electromagnetic interference signal is input, and the rectangular pulse signal can be used for carrying out transistor degradation simulation after that. Then, the rectangular pulse signal can be used as an input boundary condition and input into preset simulation software, and simulation is carried out for preset times, so that the degradation rate of the transistor to be simulated is obtained. It can be understood that the degradation effect of the rectangular pulse signal is consistent with the degradation effect of the electromagnetic interference signal, so that the degradation rate obtained after simulation according to the rectangular pulse signal can be closer to the real degradation rate, and the reliability obtained by simulation can more accurately reflect the reliability of the transistor to be simulated. Compared with the electromagnetic interference signals with more simulation points, the rectangular pulse signals can be simulated by using fewer simulation points, so that the simulation time can be reduced, and the reliability of the transistor to be simulated can be accurately and efficiently estimated.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 2 is a schematic diagram of a simulation device of some embodiments of the present application;
FIG. 3 is a schematic view of a usage scenario of a reliability simulation method of some embodiments of the present application;
FIG. 4 is a schematic diagram of pulse signal transitions for a reliability simulation method of certain embodiments of the present application;
FIG. 5 is a schematic diagram of pulse signal simulation of a reliability simulation method of certain embodiments of the present application;
FIG. 6 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 7 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 8 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 9 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 10 is a flow diagram of a reliability simulation method of certain embodiments of the present application;
FIG. 11 is a schematic diagram of pulse data and schematic diagram of probability density data for a reliability simulation method of certain embodiments of the present application;
FIG. 12 is a flow diagram of a reliability simulation method of some embodiments of the present application;
FIG. 13 is a flow diagram of a reliability simulation method of certain embodiments of the present application;
FIG. 14 is a drain current versus schematic and degradation rate versus schematic diagram of a reliability simulation method of certain embodiments of the present application;
FIG. 15 is a block diagram of a reliability simulation device of some embodiments of the present application;
fig. 16 is a schematic diagram of a connection state of a non-volatile computer readable storage medium and a processor according to some embodiments of the present application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only for explaining the embodiments of the present application and are not to be construed as limiting the embodiments of the present application.
Currently, a power transistor can be widely used as a semiconductor device in a circuit of an industrial chip. When transistors are used in circuits, workers also need to systematically optimize the transistor structure for different chip application environments to enhance the robustness of the transistor in certain scenarios. Therefore, before actually streaming a chip (i.e., manufacturing a chip through a series of process steps like a pipeline), a worker needs to perform reliability simulation on a transistor using simulation software, such as semiconductor process simulation and device simulation software (Technology Computer Aided Design, TCAD), to evaluate the lifetime of the transistor based on the calculation result.
The electromagnetic interference signal components of the common industrial chip application environment are complex, and the electromagnetic interference signal is difficult to be characterized into simulation conditions to be input into simulation software, or the reliability simulation is caused to take a long time, so that a severe challenge is brought to the real transistor degradation simulation. Prior art solutions are directed to transistors that rely mainly on signal parameters of relatively simple pulses as transient simulation inputs in reliability evaluation of electromagnetic fields, for example for laterally diffused metal oxide semiconductor (Laterally Diffused Metal Oxide Semiconductor, LDMOS) transistors, currently typically rely on signal parameters of periodic single transmission line pulses (Transmission Line Pulse, TLP) as transient simulation inputs. However, due to the simple simulation input condition of the simulation method, the periodic single transmission line pulse cannot well reflect the real electromagnetic interference signal, so that the current simulation method is difficult to cover the common complex electromagnetic pulse environment of the industrial chip, the difference between the reliability life expected value and the actual experience value of the transistor is huge, and the stability of the chip is difficult to be accurately evaluated.
In order to solve the technical problems, the embodiment of the application provides a reliability simulation method.
The reliability simulation method of the present application will be described in detail below:
referring to fig. 1 and 2, an embodiment of the present application provides a reliability simulation method, including:
step 01: acquiring a periodic interference signal of an electromagnetic interference signal, wherein the electromagnetic interference signal is formed by superposing a plurality of electromagnetic pulse signals;
specifically, the reliability simulation method of the present application may be applied to the simulation device 100, where the simulation device 100 includes a processor 30, where the processor 30 may be provided with simulation software 20, such as commercial TCAD software, and the simulation software 20 may be used to perform reliability simulation on a transistor to be simulated, so as to perform life assessment on the transistor and the chip according to a simulation result. The transistor to be simulated can be a diode or a triode, for example, the transistor to be simulated is an LDMOS transistor, and the LDMOS transistor has the advantages of low distortion, high voltage resistance, extremely high output and the like, and is widely applied to a circuit structure of an industrial chip as a power semiconductor representative device.
The devices in an industrial chip application environment are so numerous that there may be multiple pulse signals acting on the chip and transistors therein. Taking an electromagnetic interference source common to the power industry as an example, when a plurality of GAS insulated SWITCHGEAR, GIS switch devices operate in parallel at the same frequency, each GAS insulated SWITCHGEAR, GIS switch device may generate an electromagnetic pulse signal, so that a plurality of electromagnetic pulse signals are superimposed to generate an electromagnetic interference signal with complex components. In order to ensure the simulation effect, the processor 30 needs to determine the simulation input condition according to the electromagnetic interference signal actually applied on the chip, so as to ensure that the simulation result can be closer to the actual degradation effect. Therefore, the processor 30 needs to acquire the electromagnetic interference signal first, taking a common electromagnetic interference source in the power industry as an example, the processor 30 can acquire the electromagnetic interference signal at the pin of the industrial chip, and at this time, a plurality of electromagnetic pulse signals are generated by a plurality of gas-insulated switchgear devices running in parallel at the same frequency. Fig. 3 (a) is a time domain diagram of an electromagnetic pulse signal caused by a high-frequency switch arranged on a gas-insulated switch, and voltage interference generated on a chip pin when a plurality of devices run in parallel at the same frequency can be seen from the diagram.
A single electromagnetic pulse signal may form extremely short rising/falling edges and have a relatively high amplitude, so that the single electromagnetic pulse signal conforms to a gaussian distribution law in the time domain. Thus, according to the voltage superposition theorem, the processor 30 may perform gaussian mixture simulation on the electromagnetic interference signal to quantitatively analyze the electromagnetic interference signal. Meanwhile, the electromagnetic interference signals have periodicity, and the signal intensities in different periods are maintained relatively consistent as a whole, and it can be understood that the time domain functions of the Gaussian mixture interference signals corresponding to the periodic interference signals in different periods are approximately the same. The processor 30 may obtain the periodic interference signal of the electromagnetic interference signal according to the signal characteristic period of the electromagnetic interference signal, for example, fig. 3 (b), so as to perform the gaussian mixture simulation according to the periodic interference signal, without performing the gaussian mixture simulation on the whole electromagnetic interference signal.
Step 02: performing Gaussian mixture simulation on the periodic interference signal to obtain a Gaussian mixture interference signal time domain function;
specifically, after the periodic interference signal is obtained, the processor 30 may perform gaussian mixture simulation on the periodic interference signal to obtain a gaussian mixture interference signal time domain function, and obtain a distribution curve of the periodic interference signal, for example, fig. 3 (c) is a schematic diagram of a fitting process of the periodic interference signal, a light-color curve is a fitting line segment generated in the fitting process, and fig. 3 (d) is a distribution curve of the periodic interference signal obtained by fitting. Thus, the processor 30 completes the description of the periodic interference signal, thereby completing the description of the electromagnetic interference signal, so that the reliability simulation can be performed based on the time domain function of the Gaussian mixture interference signal, and further, the simulation effect is ensured to be closer to the real degradation effect.
Step 03: converting the periodic interference signal into a rectangular pulse signal according to a Gaussian mixture interference signal time domain function and a preset acceleration factor model, wherein the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the preset acceleration stress condition, and meets the degradation consistency condition, and the degradation consistency condition is the degradation consistency of the dynamic stress in the preset time period and the generation of the effective electric field static stress and the dynamic stress obtained through integration;
specifically, the reliability dynamic and static stress conversion theory is applied to convert the complex periodic interference signal into a single pulse signal (such as a rectangular pulse signal) so as to reduce the simulation complexity. The voltage of the periodic interference signal is not stable, so that the stress generated by the periodic interference signal can be regarded as dynamic stress; the voltage of the rectangular pulse signal is relatively stable, so that the stress generated by the rectangular pulse signal can be regarded as effective electric field static stress.
The conversion principle of the reliability dynamic and static stress conversion theory is as follows:
the precondition for the reliable dynamic and static stress conversion is that the stress variation amplitude is relatively small, so that the dynamic stress only accelerates the original failure mechanism, but the original failure mechanism is not changed, and the stress variation amplitude is generally small in the simulation process, so that the reliable dynamic and static stress conversion can be directly carried out in the common simulation process.
Effective electric field static stress epsilon eff The time increment d t Time increment d under' and true electric field stress ε (t) (acceleration or deceleration conditions) t The relationship between can be written asWherein->The acceleration factor model is defined as the ratio between the expected failure time under normal working conditions and the actual failure time under the condition of preset acceleration stress. Meanwhile, the preset acceleration factor model meets the degradation consistency condition, wherein the degradation consistency condition is the effective electric field static stress epsilon obtained by integration for the dynamic stress within a preset time period eff The degradation produced is consistent with that produced by dynamic stress. The preset duration may be equal to a signal characteristic period of the electromagnetic interference signal, or the preset duration may be equal to a duration corresponding to a portion of the electromagnetic interference signal, where the voltage strength is not 0, in a single signal characteristic period of the electromagnetic interference signal.
It can be understood that after the reliable dynamic and static stress conversion is performed according to the time domain function of the Gaussian mixture interference signal and the preset acceleration factor model, the degradation effect of the obtained rectangular pulse signal on the transistor to be simulated is consistent with the degradation effect of the periodic interference signal on the transistor to be simulated, so that the transistor degradation simulation can be performed by using the simpler rectangular pulse signal. For example, the solid line in fig. 4 is a periodic interference signal, the dotted line is a rectangular pulse signal, the amplitude of the rectangular pulse signal obtained after conversion is 11.64V, and the pulse width is 250ns.
Meanwhile, please refer to fig. 5, wherein the solid line in fig. 5 is a pulse signal, the points are simulation points, and it can be understood that the more simulation points, the longer the simulation time, fig. 5 (a) is a simulation point node of the reliability simulation according to a certain periodic interference signal in the electromagnetic interference signals, and fig. 5 (b) is a simulation point node of the reliability simulation according to a rectangular pulse signal. As can be seen from fig. 5 (a), for the periodic interference signal, the voltage of the periodic interference signal is continuously changed, so that the simulation points when the periodic interference signal is simulated need to be set more, so that the simulation effect can be ensured, but this results in a longer simulation time. As can be seen from fig. 5 (b), since the rectangular pulse signal is more stable, the simulation software 20 can set the number of simulation points to be smaller for the period of voltage stability when simulating the rectangular pulse signal, and can have more simulation points in the case of abrupt voltage change so as to extract more details of the rectangular pulse signal under the voltage, so that on one hand, the simulation software 20 can extract more perfect details of the rectangular pulse signal, so that the simulation software 20 can simulate the rectangular pulse signal well, and on the other hand, the simulation time can be reduced. Therefore, for the rectangular pulse signal and the periodic disturbance signal of the same period, the number of simulation points when the rectangular pulse signal is simulated is smaller than that when the periodic disturbance signal is simulated, so that the simulation time when the rectangular pulse signal is simulated is shorter than that when the electromagnetic disturbance signal is simulated.
For example, the reliability simulation of the LDMOS transistor is carried out by adopting a transient solution mode, the initial step size is 0.0001ns, the growth rate is 1.2, and the maximum step size is 100ns. For a periodic interference signal with a period of 300ns, the number of simulation points of a signal sample with a resolution of 1ns in a simulation process is 302, and the simulation time is 26 minutes. For a rectangular pulse signal, the number of simulation points in a 300ns period is 103, and the simulation time length is 8min16s. The comparison shows that the simulation duration of the device reliability after the equivalent replacement is shortened by 2/3.
In summary, the simulation effect of the reliability simulation based on the periodic disturbance signal and the reliability simulation based on the rectangular pulse signal are identical, but the simulation time of the rectangular pulse signal is shorter than that of the periodic disturbance signal. Therefore, after the time domain function of the gaussian mixture interference signal is obtained, the processor 30 performs reliability dynamic and static stress conversion on the electromagnetic interference signal collected by the experiment according to the time domain function of the gaussian mixture interference signal and a preset acceleration factor model to obtain a simpler rectangular pulse signal, so that on one hand, the signal input parameters can be simplified, the complexity of simulation is reduced, and on the other hand, the simulation time can be greatly shortened on the premise of maintaining the relative consistency of the simulation precision, and the high-efficiency and accurate simulation of the reliability can be realized.
Step 04: the rectangular pulse signal is input as an input boundary condition to a preset simulation software 20 for simulation a preset number of times to output a degradation rate of the transistor to be simulated, and the degradation rate is configured to characterize the reliability of the transistor to be simulated.
Specifically, after determining the rectangular pulse signal, the processor 30 may input the rectangular pulse signal as an input boundary condition to the preset simulation software 20, and the simulation software 20 may simulate the rectangular pulse signal to perform a simulation for a preset number of times, for example, 2000 times, using the rectangular pulse signal obtained by the simulation. The processor 30 may then determine the degradation rate of the transistor to be simulated after performing the simulation for the preset number of times according to certain parameters, for example, the transistor to be simulated includes four electrodes, namely, a Gate (Gate), a Source (Source), a Drain (Drain), and a Substrate (Substrate), and the processor 30 may determine the degradation rate of the transistor to be simulated according to the Gate voltage or the Drain current of the transistor to be simulated.
The degradation rate is configured to characterize the reliability of the transistor to be emulated. For example, the processor 30 may acquire a drain current (i.e., an initial drain current) once before the simulation software 20 performs the reliability simulation, acquire a drain current (i.e., a current drain current) again after the simulation software 20 completes the reliability simulation for a preset number of times, and then determine the degradation rate of the transistor to be simulated according to the initial drain current and the current drain current. In some embodiments, the degradation rate is characterized by a first ratio of the present drain current and the initial drain current, the greater the present drain current, the greater the first ratio, which represents less transistor degradation, the smaller the degradation rate, and the higher the reliability. In other embodiments, the degradation rate is characterized by a second ratio of the difference between the initial drain current and the present drain current to the initial drain current, the greater the present drain current, the smaller the difference between the initial drain current and the present drain current, the smaller the second ratio, which is indicative of less transistor degradation, the smaller the degradation rate and the greater the reliability.
The reliability simulation method of the embodiment of the application firstly acquires an electromagnetic interference signal formed by superposition of a plurality of electromagnetic pulse signals, and then preprocesses the electromagnetic interference signal to acquire a periodic interference signal of the electromagnetic interference signal. And then, carrying out Gaussian mixture simulation on the periodic interference signal, and obtaining a Gaussian mixture interference signal time domain function corresponding to the periodic interference signal after simulation so as to describe the complex electromagnetic interference signal, so that the follow-up reliability simulation can be carried out based on the Gaussian mixture interference signal time domain function, and the simulation effect is ensured to be closer to the actual degradation effect. Then, an acceleration factor model meeting the degradation consistency condition can be preset, and the periodic interference signal is converted according to the preset acceleration factor model to obtain a rectangular pulse signal. At this time, the stress generated by the periodic disturbance signal may be regarded as dynamic stress, and the stress generated by the converted rectangular pulse signal may be regarded as effective electric field static stress. The rectangular pulse signal retains the rising and falling characteristics of the signal, the degradation generated by the static stress and the dynamic stress of the effective electric field is consistent, and the characteristic that the periodic interference signal can represent the electromagnetic interference signal is combined, so that it can be understood that when the rectangular pulse signal is input to the simulation software 20 and the electromagnetic interference signal is input, the degradation condition of the transistor is consistent, and therefore the rectangular pulse signal can be used for the transistor degradation simulation subsequently. Then, the rectangular pulse signal can be used as an input boundary condition and input into the preset simulation software 20, and the simulation is performed for preset times to obtain the degradation rate of the transistor to be simulated. It can be understood that the degradation effect of the rectangular pulse signal is consistent with the degradation effect of the electromagnetic interference signal, so that the degradation rate obtained after simulation according to the rectangular pulse signal can be closer to the real degradation rate, and the reliability obtained by simulation can more accurately reflect the reliability of the transistor to be simulated. Compared with the electromagnetic interference signals with more simulation points, the rectangular pulse signals can be simulated by using fewer simulation points, so that the simulation time can be reduced, and the reliability of the transistor to be simulated can be accurately and efficiently estimated.
Referring to fig. 2 and 6, in some embodiments, step 03: according to the Gaussian mixture interference signal time domain function and a preset acceleration factor model, converting the periodic interference signal into a rectangular pulse signal, wherein the method comprises the following steps:
step 031: determining a preset acceleration factor model according to preset stress parameters of the transistor to be simulated;
step 032: acquiring equivalent pulse voltage according to a degradation consistency condition, a preset acceleration factor model, a Gaussian mixture interference signal time domain function, a gate oxide thickness of a transistor to be simulated, an effective starting time of a periodic interference signal and an effective ending time of the periodic interference signal;
step 033: and determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
Specifically, the preset stress parameter is a parameter related to an electric field in a preset failure time model, and the preset stress parameter reflects the influence of different electric field stresses on the degradation of the transistor to be simulated, so that it can be understood that the preset stress parameter is essential in the reliability dynamic and static stress conversion.
After obtaining the preset stress parameters, the processor 30 may determine a preset acceleration factor model according to the preset stress parameters and failure mechanism of the transistor to be simulated, and then the processor 30 may perform signal conversion according to the preset acceleration factor model and degradation consistency conditions.
For example, fig. 4 shows a periodic disturbance signal and a rectangular pulse signal corresponding to the LDMOS transistor. The transistor reliability simulation is based on the simulation of defect theory, and aims at the LDMOS transistor, interface state defects caused by the breakage of Si-H bonds of a silicon-oxygen interface inside the transistor have bad influence on the degradation of the transistor. At this time, the failure time model usually adopts an exponential form to describe the field effect, and the acceleration factor model corresponding to the LDMOS transistor is as follows
Wherein ε (t) is the true electric field stress, ε eff To be effective electric field static stress, effective electric field static stress epsilon eff And the true electric field stress ε (t) in this case represent the effective electric field strength and the true electric field strength, respectively.
The degenerate consistency condition may include:
wherein P is the effective duration of the signal, i.e. the time range corresponding to the part of the signal where the voltage strength is not 0, and t is time.
By combining the formula (1) and the formula (2), and converting the electric field strength into a voltage (for example, after multiplying the electric field strength by the gate oxide thickness of the transistor to be simulated, a corresponding voltage can be obtained), the formula can be obtained:
wherein, gamma is a preset stress parameter,for equivalent pulse voltage, ta is the effective start time of the periodic disturbance signal, tb is the effective stop time of the periodic disturbance signal, tox is the gate oxide thickness of the transistor to be simulated, and the gate oxide thickness of the LDMOS transistor is 38nm, as can be seen from FIG. 4 a 30ns, t b 280ns, and the pulse width was 250ns. When processor 30 simulates an LDMOS transistor, the equivalent pulse voltage can be obtained by taking all data>At 11.64V, the processor 30 then obtains the rectangular pulse signal of fig. 4.
Therefore, after obtaining the predetermined acceleration factor model, the processor 30 may combine the degradation consistency condition with the predetermined acceleration factor model, and then obtain the equivalent pulse voltage according to the gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective start time of the periodic interference signal, and the effective end time of the periodic interference signal. The effective start time of the periodic interference signal is a time when the voltage of the periodic interference signal starts to rise from 0 or a value very close to 0, and the effective end time of the periodic interference signal is a time when the voltage of the periodic interference signal will be 0 or a value very close to 0. The effective duration of the periodic interference signal is a time range corresponding to a part of the periodic interference signal, the voltage strength of which is not 0, namely a time range between the effective starting time and the effective ending time of the periodic interference signal.
The effective duration before and after signal conversion is not changed, the effective starting time of the periodic interference signal is the same as the effective starting time of the rectangular pulse signal, and the effective ending time of the periodic interference signal is the same as the effective ending time of the rectangular pulse signal. The processor 30 may determine the effective start time of the rectangular pulse signal based on the effective start time of the periodic disturbance signal and the effective end time of the rectangular pulse signal based on the effective end time of the periodic disturbance signal, thereby determining the waveform and pulse width of the rectangular pulse signal. For example, the effective start time in fig. 4 is 30ns, the effective end time is 280ns, and the pulse width is 250ns.
In this way, the processor 30 accurately acquires the equivalent pulse voltage according to the simultaneous preset acceleration factor model and degradation consistency condition, and according to the gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective start time of the periodic interference signal and the effective end time of the periodic interference signal required to be used in the simultaneous formula. In combination with the equivalent pulse voltage, the effective start time and the effective end time, the processor 30 may generate a rectangular pulse signal having a degradation effect consistent with the degradation effect of the periodic disturbance signal after conversion.
Referring to fig. 2 and 7, in some embodiments, step 031: determining a preset acceleration factor model according to preset stress parameters of a transistor to be simulated, including:
step 0311: performing an accelerating electric field experiment on a plurality of test transistors at constant temperature, and obtaining the actual failure time of each test transistor by applying constant electric fields with different intensities to each test transistor, wherein the specification and the model of the plurality of test transistors are the same;
step 0312: taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated;
Step 0313: acquiring preset stress parameters according to the failure time model and a group of data; and
Step 0314: and acquiring a preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
Specifically, the processor 30 may accelerate the failure of the transistor to be simulated by strengthening test conditions without changing the failure mechanism of the transistor, so as to obtain necessary information, such as preset stress parameters, in a shorter time to evaluate the life index of the device under normal conditions.
First, the processor 30 performs an accelerated electric field experiment on a plurality of test transistors with the same specification and model at constant temperature, and the processor 30 obtains the actual failure time of each test transistor by applying constant electric fields with different intensities. And then taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the types of the test transistors and the transistors to be simulated are the same so as to ensure the accuracy of the obtained preset stress parameters and the accuracy of the rectangular pulse signals after subsequent conversion. After fitting to obtain the dead time model, the processor 30 may calculate the preset stress parameter according to the dead time model and a set of data.
For example, the test transistor is an LDMOS transistor, and the experimental conditions of the accelerated electric field test of the LDMOS transistor are simple and short in time consumption, and the exponential failure time model tf=a can be obtained by applying constant electric fields with different intensities at the device end 0 exp(-γε)exp(Q/k B T), wherein TF is the device failure time, A 0 Is the coefficient of the process-related model factor, gamma and Q are stress parameters in an exponential relationship, wherein gamma is the coefficient related to an electric field (namely preset stress parameters), Q is the coefficient related to temperature, epsilon is the stress field applied, and k B Is the boltzmann constant, and T is the temperature. After calculation, the stress parameter gamma= 2.663cm/MV is obtained by an LDMOS transistor accelerating electric field experiment.
In combination with the above formula (1), the processor 30 may obtain the preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress obtained by the simple electric field acceleration experiment.
In this way, the processor 30 can determine the preset acceleration factor model by combining the fitted gaussian mixture interference signal time domain function with the preset stress parameter obtained by the constant electric field experiment with short time consumption, so as to equivalent the complex electromagnetic interference signal to a simple rectangular pulse signal, thereby greatly shortening the simulation time and reducing the complexity of the simulation. And the processor 30 can calibrate the acquired data by using the experimental result (i.e. the preset stress parameter obtained by the experiment), thereby improving the reliability of the simulation. Meanwhile, the time consumption of the device acceleration experiment under the constant electric field is short, so that the reliability simulation method can greatly shorten the reliability simulation time.
Referring to fig. 2 and 8, in some embodiments, step 01: acquiring a periodic interference signal of the electromagnetic interference signal, including:
step 011: acquiring a signal characteristic period of an electromagnetic interference signal;
step 012: and acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
Specifically, the peak pulse value of each pulse signal of the electromagnetic interference signal is slightly floating under the influence of environmental factors, i.e. the peak pulse value of each pulse signal may be different, but the difference is not large, and the relative uniformity is maintained as a whole.
After the electromagnetic interference signal is obtained, the processor 30 may perform preprocessing on the electromagnetic interference signal, that is, sampling and extracting the electromagnetic interference signal to obtain a signal characteristic period of the electromagnetic interference signal, and may also obtain a peak voltage of the electromagnetic interference signal at this time. Processor 30 then obtains a periodic interference signal from the electromagnetic interference signal based on the signal characteristic period. For example, the periodic interference signal shown in fig. 3 (b) is obtained from the electromagnetic interference signal shown in fig. 3 (a), and the characteristic period of the electromagnetic interference signal is 300ns, the peak value is 16.1V, and the side peak values are 10.2V and 12.1V, respectively. In this way, since the periodic interference signal is substantially the same as other interference signals in the electromagnetic interference signal, the gaussian mixture interference signal time domain function determined according to the periodic interference signal can represent the gaussian mixture interference signal time domain function of the electromagnetic interference signal, so that the processor 30 can complete the description of the electromagnetic interference signal for the periodic interference signal, thereby simplifying the gaussian simulation process and accelerating the gaussian simulation speed.
Referring to fig. 2 and 9, in some embodiments, step 02: performing Gaussian mixture simulation on the periodic interference signal to obtain a Gaussian mixture interference signal time domain function, wherein the Gaussian mixture simulation comprises the following steps:
step 021: acquiring a target Gaussian mixture model corresponding to the wave crest number according to the wave crest number of the periodic interference signal;
step 022: and carrying out Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm so as to obtain a Gaussian mixture interference signal time domain function.
Specifically, the probability function of the Gaussian mixture model isRepresenting a linear combination of K gaussian distributions, the overall parameter is Θ= { α 1 ,K,α K ,μ 1 ,K,μ K ,σ 1 ,K,σ K }. Wherein alpha is i Is the ith Gaussian distribution probability weighting value and has +.>Represents the probability density of x under the ith gaussian distribution,represents the ith Gaussian distribution, μ i For the distribution mean value, sigma i Is the standard deviation. According to the voltage superposition theorem, the processor 30 obtains each gaussian distribution function, and performs weighted summation on the obtained gaussian distribution functions to obtain a time domain function of the gaussian mixed interference signal, where x in the above formula refers to time in the present application because the time domain function is obtained.
The processor 30 may determine the number of K in the target gaussian mixture model based on the number of peaks of the periodic interference signal, for example, three peaks in fig. 3 (b), at which point the processor 30 may determine that the target gaussian mixture model is composed of three gaussian distribution functions, and thus may determine that the target gaussian mixture model is a 3-state gaussian mixture model, i.e., that the K value in the target gaussian mixture model is 3.
The expectation maximization (ExpectationMaximization, EM) algorithm is an algorithm that looks for a parameter maximum likelihood estimate or a maximum a posteriori estimate in a probabilistic model. Wherein the formula of the maximum likelihood function isThe processor 30 is acquiring the distribution data x= { X 1 ,x 2 ,K,x N After the process of the Gaussian mixture model parameter Θ simulation is carried out on the distributed data, the processor 30 constructs the maximum likelihood function, then calculates the likelihood function value based on the initial value of the parameter, adjusts the parameter according to the likelihood function value, then calculates the likelihood function value according to the adjusted parameter value, and the parameter with the maximum likelihood function value after repeated iterative calculation can be regarded as the optimal parameter->The posterior probability is a value indicating the probability that the cause of occurrence of the event is caused by a factor under the condition that the event has occurred, for example, the posterior probability is expressed as +.>Where k=1, 2,..n, i=1, 2,..k. At this time, the processor 30 calculates the posterior probability based on the initial value of the parameter, adjusts the parameter according to the posterior probability, and calculates the likelihood function value according to the adjusted parameter value, so that the parameter with the maximum posterior probability can be considered as the optimal parameter +.>Alternatively, in the maximum likelihood estimation and in the maximum a posteriori estimation, when the difference between the two iterations is smaller than the error value, the parameter at this time can be regarded as the optimal parameter +. >
The processor 30 performs gaussian mixture simulation on the periodic interference signal according to the target gaussian mixture model by adopting an expected maximum algorithm, and the obtained optimal parameter is the parameter in the time domain function of the gaussian mixture interference signal. After combining the optimal parameters obtained by the desired maximum algorithm and the target gaussian mixture model, the processor 30 may obtain a gaussian mixture interference signal time domain function corresponding to the periodic interference signal. It will be appreciated that the resulting gaussian mixture of interference signal time domain functions also corresponds to the whole electromagnetic interference signal.
In this way, the processor 30 may determine a target gaussian mixture model suitable for the periodic interference signal by using the number of peaks of the periodic interference signal, and acquire the optimal parameters of the target gaussian mixture model by using the expected maximum algorithm, so as to determine the gaussian mixture interference signal time domain function, so that the gaussian mixture interference signal time domain function can be more consistent with the periodic interference signal, thereby improving the accuracy of the subsequent reliability simulation.
Referring to fig. 2 and 10, in certain embodiments, step 022: performing Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm to obtain a Gaussian mixture interference signal time domain function, wherein the Gaussian mixture simulation comprises the following steps:
Step 0221: converting the voltage intensity in the time domain of the periodic interference signal into probability density in the time domain, wherein the probability density is determined according to the ratio of the voltage intensity value to the total voltage intensity value;
step 0222: according to preset distribution data, obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the Gaussian mixture model, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight values, distribution mean values and standard deviations of each Gaussian distribution function;
step 0223: and generating a Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
Specifically, when the gaussian mixture simulation is performed on the periodic interference signal based on the expected maximum value algorithm, the processor 30 first needs to convert the voltage intensity in the time domain of the periodic interference signal into a probability density in the time domain, the processor 30 may perform nanosecond sampling extraction on the pulse data of the periodic interference signal, and accumulate to obtain a total voltage intensity value, where the probability density in the time domain may be calculated by the formula probability density=time domain voltage value/total voltage intensity value, for example, fig. 11 (a) shows the pulse data of the periodic interference signal, and fig. 11 (b) shows the probability density in the time domain.
Processor 30 then determines the number of gaussian distribution functions in the gaussian mixture model probability function from the target gaussian mixture model. For example, if the target gaussian mixture model is a 3-state gaussian mixture model, the number of gaussian distribution functions in the gaussian mixture model probability functions corresponding to the target gaussian mixture model can be regarded as 3. The processor 30 may then randomly acquire the parameter values of the gaussian distribution functions, i.e., randomly acquire the weight, the distribution mean, and the standard deviation of each gaussian distribution function, to weight sum based on one or more gaussian distribution functions to obtain a gaussian mixture model probability function.
The periodic disturbance signal characterizes a trend of change of the voltage over time, so the preset distribution data may be a set of simulation times. After acquiring the gaussian mixture model probability function, the processor 30 may calculate an optimal parameter of the gaussian mixture model probability function according to the preset distribution data, the probability density corresponding to the preset distribution data, and the expected maximum value algorithm, and then adjust the parameter of the gaussian mixture model probability function according to the optimal parameter, thereby generating the gaussian mixture interference signal time domain function. It can be understood that the time domain function of the Gaussian mixture interference signal generated at this time has higher accuracy in describing the periodic interference signal, so that higher simulation accuracy in the subsequent simulation based on the time domain function of the Gaussian mixture interference signal is ensured.
In this way, the processor 30 can disassemble and analyze the periodic interference signal based on the expected maximum algorithm to obtain the gaussian mixture interference signal time domain function, so as to accurately characterize the complex periodic interference signal.
Referring to fig. 2 and 12, in certain embodiments, step 0222: according to preset distribution data, calculating optimal parameters of a Gaussian mixture model probability function corresponding to the Gaussian mixture model, wherein the optimal parameters comprise:
step 02221: different parameter combinations are obtained, wherein the parameter combinations comprise weight values, distribution mean values and standard deviations of each Gaussian distribution function;
step 02222: calculating the distribution probability of preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability;
step 02223: and determining a target parameter combination, wherein the target parameter combination is a parameter combination with an error value smaller than a preset error value of the parameter combination before and after updating.
Specifically, after determining the number of gaussian distribution functions, the processor 30 randomly obtains a parameter combination of each gaussian distribution function, where the parameter combination includes a weight, a distribution mean, and a standard deviation, and the sum of the weights of each gaussian distribution function is 1.
The processor 30 calculates the parameters at eachUnder the combination, the distribution probability of each preset distribution data belonging to the Gaussian distribution function corresponding to the parameter combination is set, and the parameter combination is updated according to the distribution probability. For example, according to the formulaWhere k=1, 2..n, i=1, 2..k, and calculating to obtain the distribution probability of each preset distribution data from the gaussian distribution function corresponding to a certain parameter combination. Then-> Is->Updating parameters in the parameter combination. After updating the parameters, the processor 30 calculates the distribution probability of each preset distribution data from the gaussian distribution function corresponding to a certain parameter combination again, then updates the parameters according to the calculated distribution probability again, and repeats the steps of calculating the distribution probability and updating the parameters until the parameter combination difference (i.e. the error value of the parameter combination before and after updating) between the two iterations is smaller than the preset error value, so that the fitting can be considered to be completed. The parameter combination at this time is the target parameter combination, and the processor 30 can adjust the parameter combination of the corresponding gaussian distribution function according to the target parameter combination, generate a gaussian mixture model probability function after parameter adjustment, and further generate a gaussian mixture interference signal time domain function according to the gaussian mixture model probability function after parameter adjustment.
After determining the target parameter combination, the processor 30 may adjust parameters of the gaussian mixture model probability function according to the target parameter combination, and then convert the gaussian mixture model probability function after parameter adjustment into an electrical signal, so as to complete signal analysis of the periodic interference signal, and obtain pulse data of the periodic interference signal, thereby facilitating subsequent implementation of reliable dynamic and static stress conversion according to the pulse data. For example, the processor 30 may calculate the following formula: and obtaining pulse data of the periodic electromagnetic interference signals corresponding to each preset distribution data by using the probability density of the periodic electromagnetic signals = Gaussian mixture model probability function and the total voltage intensity value.
In addition, the processor 30 also multiplies each gaussian distribution function by the voltage intensity total value and then performs weighted summation on the products to obtain a gaussian mixture interference signal time domain function, for example, the gaussian mixture interference signal time domain function isWherein V (t) is a Gaussian mixture interference signal time domain function, t is time, and more than five hundred numerical values in each Gaussian distribution function are products of the total voltage intensity value and the weight of the corresponding Gaussian distribution function. In this way, the processor 30 can improve the correlation between the gaussian mixture interference signal time domain function and the pulse data, and avoid that the data obtained by the gaussian mixture interference signal time domain function is too small, which results in subsequent calculation difficulty.
In this way, through continuous iterative computation, the processor 30 can obtain the optimal parameter combination of each gaussian distribution function, so that the probability function of the electromagnetic interference signal can be more accurately described by the gaussian mixed interference signal time domain function obtained according to the gaussian distribution function.
Referring to fig. 2 and 13, in certain embodiments, step 04: inputting the rectangular pulse signal as an input boundary condition to the preset simulation software 20 for simulation for a preset number of times to output the degradation rate of the transistor to be simulated, including:
step 041: determining a plurality of simulation points in the rectangular pulse signal;
step 042: sequentially inputting voltages corresponding to a plurality of simulation points as gate voltages of a transistor to be simulated, and performing simulation for preset times, wherein the simulation is one-time when the input of the plurality of simulation points of the rectangular pulse signal is completed;
step 043: detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
Specifically, before performing the degradation simulation, the processor 30 may perform a quasi-static simulation on the transistor to be simulated, which is not performed with the degradation simulation, to determine the initial drain current, where the quasi-static simulation is a detection method that may obtain an electrical characteristic curve of the transistor (e.g., may obtain a characteristic curve of the drain voltage and the drain current), and is considered to be static during the performance of the quasi-static simulation, and does not cause degradation of the transistor, so the quasi-static simulation is generally used for electrical test simulation.
When a rectangular pulse signal is simulated, the pulse device determines a plurality of simulation points based on the rectangular pulse signal, then sequentially inputs voltages corresponding to the simulation points as gate voltages of the transistor to be simulated, and performs simulation for preset times, wherein only when the input of the simulation points of the rectangular pulse signal is completed, the simulation can be considered to be completed once, so that the transistor to be simulated is degraded. After the predetermined number of simulations, the processor 30 may perform a quasi-static simulation on the transistor to be simulated, which has been degraded, to detect the present drain current of the transistor to be simulated. Processor 30 may then determine the degradation rate based on the present drain current and the initial drain current to evaluate the reliability of the transistor to be emulated.
Referring to fig. 14 (a), the horizontal axis represents the drain voltage (Vd), the vertical axis represents the drain current (Id), the data in the diagram represents the data obtained by performing quasi-static simulation in the on state (gate voltage (Vg) =5v), the curve with five-pointed star symbols represents the initial drain current, the curve with triangle symbols represents the electromagnetic interference signal as the boundary condition, the current drain current obtained by performing 2000 simulation cycles, the curve with circular symbols represents the rectangular pulse signal as the boundary condition, and the current drain current obtained by performing 2000 simulation cycles. It can be seen that after 2000 degradation cycles, both the simulation with the electromagnetic interference signal as the boundary condition and the simulation with the rectangular pulse signal as the boundary condition, the corresponding present drain current is reduced relative to the initial drain current, so that it can be demonstrated that the transistor to be simulated is degraded after both simulations are performed.
The system of fig. 14 (b) organizes the operational state current simulation degradation conditions (vg=5v, vd=20v) of the saturation region of the LDMOS transistor. In the figure, the horizontal axis represents the number of cycles (Cycle), and the vertical axis represents the degradation rate (Δid). For effective comparison, a conventional TLP simulation model is also introduced for reference, and the conventional TLP simulation model usually uses a Full-width half maximum (FWHM) method (Full-width at the half of the maximum, FWHM) to process the complex electromagnetic signal, where the pulse width is 135ns and the pulse amplitude is 16.1V. The curve with five-pointed star marks in the figure represents the variation of the degradation rate when the traditional TLP simulation model is used for simulation, the curve with triangle marks represents the reliability simulation model taking the electromagnetic interference signal as the boundary condition, the curve with the circle marks represents the reliability simulation model taking the rectangular pulse signal as the boundary condition, and the variation of the degradation rate when the simulation is performed. The electromagnetic interference signal is directly obtained from the environment, so that degradation simulation based on the electromagnetic interference signal can be used for characterizing the degradation condition of the transistor in actual use to a large extent. Comparing the three curves, it can be known that the degradation rate of the reliability simulation model which introduces the electromagnetic interference signal or the rectangular pulse signal thereof as the simulation boundary condition is less than 1%, and the difference between the two is 0.06%. The degradation rate of the current of the reliability simulation model after 2000 stress cycles is 1.87% by using the conventional TLP simulation model, which is greatly improved compared with the degradation rate of the reliability simulation model which introduces electromagnetic interference signals as simulation boundary conditions, and the simulation life of the LDMOS transistor is usually underestimated when the simulation life of the LDMOS transistor is estimated according to the simulation result of the conventional TLP simulation model.
Therefore, as can be seen from fig. 14 (a) and 14 (b), the reliability simulation model introducing the rectangular pulse signal as the boundary condition can actually cause the degradation of the transistor to be simulated, and the accuracy of the simulation result of the reliability simulation model is high, so that the processor 30 can efficiently and accurately evaluate the lifetime of the transistor to be simulated. It can be appreciated that the present application can actually provide a more efficient and accurate reliability simulation method. In addition, the method and the device can be applied to mainstream TCAD simulation software 20 as an electromagnetic multi-physical field targeted simulation module for performing reliability simulation on the transistor to be simulated, so that the reliability simulation accuracy of the industrial chip device is improved.
Referring to fig. 2, in some embodiments, the simulation software 20 performs simulation according to a predetermined reliability simulation model, which is determined according to the type of the transistor to be simulated.
Specifically, the simulation software 20 performs simulation according to a preset reliability simulation model, and failure mechanisms of different types of transistors to be simulated may be different, and reliability simulation models corresponding to the different failure mechanisms may be different, so that the simulation software 20 also needs to determine a suitable reliability simulation model according to the type of the transistors to be simulated, so as to describe the degradation physical process of the transistors to be simulated more accurately by using the suitable reliability simulation model, thereby ensuring the simulation effect of the simulation software 20.
For example, for an LDMOS transistor, the degradation of device electrical performance is reflected in the microscopic level as an accumulation of defect states within the structure over time. Among them, the crystal bond defect growth caused by the fracture of Si-H bond on the device hetero-interface Si/SiO2 is of great importance to the influence of the device performance. Therefore, the simulation software 20 can describe the degradation mechanism of the heterojunction of the device by adopting a reactive diffusion model, and perform the reliability degradation curve simulation of the LDMOS transistor based on the degradation mechanism.
Referring to fig. 15, in order to better implement the reliability simulation method according to the embodiment of the present application, the embodiment of the present application further provides a reliability simulation apparatus 10. The reliability simulation apparatus 10 may include an acquisition module 11, an analog module 12, a conversion module 13, and a simulation module 14. The acquisition module 11 is configured to acquire a periodic interference signal of an electromagnetic interference signal, where the electromagnetic interference signal is formed by stacking a plurality of electromagnetic pulse signals. The simulation module 12 is configured to perform gaussian mixture simulation on the periodic interference signal to obtain a gaussian mixture interference signal time domain function. The conversion module 13 is configured to convert the periodic interference signal into a rectangular pulse signal according to a gaussian mixture interference signal time domain function and a preset acceleration factor model, where the preset acceleration factor model is a ratio of an expected failure time under a normal working condition to a real failure time under a preset acceleration stress condition, and the preset acceleration factor model satisfies a degradation consistency condition, where the degradation consistency condition is consistency of degradation of the dynamic stress in a preset time period and the generation of the integrated effective electric field static stress and the integrated dynamic stress. The simulation module 14 is configured to input the rectangular pulse signal as an input boundary condition to the preset simulation software 20 to simulate a preset number of times, so as to output a degradation rate of the transistor to be simulated, where the degradation rate is configured to characterize the reliability of the transistor to be simulated.
The conversion module 13 is specifically configured to determine a preset acceleration factor model according to a preset stress parameter of a transistor to be simulated; acquiring equivalent pulse voltage according to a degradation consistency condition, a preset acceleration factor model, a Gaussian mixture interference signal time domain function, a gate oxide thickness of a transistor to be simulated, an effective starting time of a periodic interference signal and an effective ending time of the periodic interference signal; and determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
The conversion module 13 is specifically configured to perform an accelerating electric field experiment on a plurality of test transistors at a constant temperature, and obtain an actual failure time of each test transistor by applying a constant electric field with different intensities to each test transistor, where the specifications and the models of the plurality of test transistors are the same; taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated; acquiring preset stress parameters according to the failure time model and a group of data; and acquiring a preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
The acquiring module 11 is specifically configured to acquire a signal characteristic period of the electromagnetic interference signal; and acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
The simulation module 12 is specifically configured to obtain a target gaussian mixture model corresponding to the number of peaks according to the number of peaks of the periodic interference signal; and carrying out Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm so as to obtain a Gaussian mixture interference signal time domain function.
The simulation module 12 is specifically configured to convert the voltage intensity in the time domain of the periodic interference signal into a probability density in the time domain, where the probability density is determined according to a ratio of the voltage intensity value to the total voltage intensity value; according to preset distribution data, obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the Gaussian mixture model, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight values, distribution mean values and standard deviations of each Gaussian distribution function; and generating a Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
The simulation module 12 is specifically configured to obtain different parameter combinations, where the parameter combinations include a weight, a distribution mean, and a standard deviation of each gaussian distribution function; calculating the distribution probability of preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability; and determining a target parameter combination, wherein the target parameter combination is a parameter combination with an error value smaller than a preset error value of the parameter combination before and after updating.
The simulation module 14 is specifically configured to determine a plurality of simulation points in the rectangular pulse signal; sequentially inputting voltages corresponding to a plurality of simulation points as gate voltages of a transistor to be simulated, and performing simulation for preset times, wherein the simulation is one-time when the input of the plurality of simulation points of the rectangular pulse signal is completed; detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
The reliability simulation device 10 is described above in connection with the accompanying drawings from the perspective of functional modules, which may be implemented in hardware, instructions in software, or a combination of hardware and software modules. Specifically, each step of the method embodiments in the embodiments of the present application may be implemented by an integrated logic circuit of hardware in the processor 30 and/or an instruction in software form, and the steps of the method disclosed in connection with the embodiments of the present application may be directly implemented as a hardware encoding processor or implemented by a combination of hardware and software modules in the encoding processor. Alternatively, the software modules may be located in a well-established storage medium in the art such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, and the like. The storage medium is located in a memory, and the processor 30 reads information in the memory and, in combination with its hardware, performs the steps of the method embodiments described above.
Referring to fig. 2, the simulation apparatus 100 of the embodiment of the present application includes a memory 40 and a processor 30, where the memory 40 stores a computer program, and when the computer program is executed by the processor 30, the processor 30 is caused to execute the reliability simulation method of any of the embodiments. For brevity, the description is omitted here.
Referring to fig. 16, the embodiment of the present application further provides a computer readable storage medium 300, on which a computer program 310 is stored, where the computer program 310, when executed by the processor 30, implements the steps of the reliability simulation method of any of the above embodiments, which is not described herein for brevity.
In the description of the present specification, reference to the terms "certain embodiments," "in one example," "illustratively," and the like, means that a particular feature, structure, material, or characteristic described in connection with the embodiments or examples is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the present application, and that variations, modifications, alternatives, and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the present application.
Claims (32)
1. A reliability simulation method, comprising:
the method comprises the steps of obtaining a periodic interference signal of an electromagnetic interference signal, wherein the electromagnetic interference signal is formed by superposing a plurality of electromagnetic pulse signals;
performing Gaussian mixture simulation on the periodic interference signal to obtain a Gaussian mixture interference signal time domain function;
Converting the periodic interference signal into a rectangular pulse signal according to the Gaussian mixture interference signal time domain function and a preset acceleration factor model, wherein the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the preset acceleration stress condition, and the preset acceleration factor model meets the degradation consistency condition, and the degradation consistency condition is that the degradation of the dynamic stress in the preset time period and the effective electric field static stress obtained through integration is consistent with the degradation of the generation of the dynamic stress; and
And taking the rectangular pulse signal as an input boundary condition, inputting the rectangular pulse signal into preset simulation software for simulation for preset times to output the degradation rate of the transistor to be simulated, wherein the degradation rate is configured to characterize the reliability of the transistor to be simulated.
2. The reliability simulation method of claim 1, wherein the converting the periodic interference signal into a rectangular pulse signal according to the gaussian mixture interference signal time domain function and a preset acceleration factor model comprises:
determining the preset acceleration factor model according to preset stress parameters of the transistor to be simulated;
acquiring equivalent pulse voltage according to degradation consistency conditions, the preset acceleration factor model, the Gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective starting moment of the periodic interference signal and the effective ending moment of the periodic interference signal; and
And determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
3. The reliability simulation method according to claim 2, wherein the determining the predetermined acceleration factor model according to the predetermined stress parameter of the transistor to be simulated comprises:
performing an accelerating electric field experiment on a plurality of test transistors at constant temperature, and obtaining the actual failure time of each test transistor by applying constant electric fields with different intensities to each test transistor, wherein the specification and the model of the plurality of test transistors are the same;
taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated;
acquiring preset stress parameters according to the failure time model and a group of data; and
And acquiring the preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
4. The reliability simulation method according to claim 1 or 2, wherein the degradation consistency condition includes:
wherein P is the preset time length,and t is time for the preset acceleration factor model, and the preset duration is determined according to the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
5. The reliability simulation method of claim 1, wherein the acquiring the periodic interference signal of the electromagnetic interference signal comprises:
acquiring a signal characteristic period of the electromagnetic interference signal; and
And acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
6. The reliability simulation method of claim 1 wherein performing gaussian mixture simulation on the periodic interference signal to obtain a gaussian mixture interference signal time domain function comprises:
acquiring a target Gaussian mixture model corresponding to the wave crest number according to the wave crest number of the periodic interference signal; and
And carrying out Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm so as to obtain a Gaussian mixture interference signal time domain function.
7. The reliability simulation method of claim 6 wherein performing gaussian mixture simulation on the periodic interference signal according to the target gaussian mixture model using a desired maximum algorithm to obtain a gaussian mixture interference signal time domain function comprises:
converting the voltage intensity in the time domain of the periodic interference signal into probability density in the time domain, wherein the probability density is determined according to the ratio of the voltage intensity value to the total voltage intensity value;
obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the target Gaussian mixture model according to preset distribution data, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight, distribution mean value and standard deviation of each Gaussian distribution function; and
And generating the Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
8. The reliability simulation method according to claim 7, wherein the calculating the optimal parameters of the gaussian mixture model probability function corresponding to the target gaussian mixture model according to the preset distribution data includes:
Different parameter combinations are obtained, wherein the parameter combinations comprise weight, distribution mean and standard deviation of each Gaussian distribution function;
calculating the distribution probability of the preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability;
and determining a target parameter combination, wherein the target parameter combination is the parameter combination with the error value of the parameter combination before and after updating being smaller than a preset error value.
9. The reliability simulation method according to claim 1, wherein inputting the rectangular pulse signal as an input boundary condition to a preset simulation software for a preset number of simulations to output a degradation rate of a transistor to be simulated comprises:
determining a plurality of simulation points in the rectangular pulse signal;
sequentially inputting voltages corresponding to the simulation points as gate voltages of the transistor to be simulated, and performing simulation for preset times, wherein the simulation is performed once when the input of the simulation points of the rectangular pulse signal is completed; and
Detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
10. The reliability simulation method of claim 9 wherein the degradation rate is characterized by a first ratio of the present drain current and the initial drain current or the degradation rate is characterized by a second ratio of a difference between the initial drain current and the present drain current to the initial drain current.
11. The reliability simulation method according to claim 1, wherein the number of simulation points when simulation is performed with the rectangular pulse signal is smaller than the number of simulation points when simulation is performed with the periodic disturbance signal.
12. The reliability simulation method according to claim 1, wherein the simulation software performs simulation according to a preset reliability simulation model, and the reliability simulation model is determined according to the type of the transistor to be simulated.
13. The reliability simulation method of claim 12 wherein the transistor to be simulated comprises a laterally diffused metal oxide semiconductor transistor and the reliability simulation model comprises a reactive diffusion model.
14. The reliability simulation method according to claim 1, wherein a plurality of the electromagnetic pulse signals are generated by running a plurality of gas-insulated switchgear devices in parallel at the same frequency, respectively.
15. The reliability simulation method according to claim 1, wherein the transistor to be simulated comprises a diode or a triode.
16. A reliability simulation device, characterized in that the reliability simulation device comprises:
the acquisition module is used for acquiring a periodic interference signal of an electromagnetic interference signal, wherein the electromagnetic interference signal is formed by superposing a plurality of electromagnetic pulse signals;
the simulation module is used for carrying out Gaussian mixture simulation on the periodic interference signal so as to obtain a Gaussian mixture interference signal time domain function;
the conversion module is used for converting the periodic interference signal into a rectangular pulse signal according to the Gaussian mixture interference signal time domain function and a preset acceleration factor model, wherein the preset acceleration factor model is the ratio of the expected failure time under the normal working condition to the actual failure time under the preset acceleration stress condition, and the preset acceleration factor model meets the degradation consistency condition, and the degradation consistency condition is that the degradation of the dynamic stress in the preset time period and the effective electric field static stress obtained by integration and the generation of the dynamic stress are consistent; and
And the simulation module is used for inputting the rectangular pulse signal as an input boundary condition to preset simulation software for simulation for preset times so as to output the degradation rate of the transistor to be simulated, wherein the degradation rate is configured to characterize the reliability of the transistor to be simulated.
17. The reliability simulation device of claim 16, wherein the conversion module is further configured to: determining the preset acceleration factor model according to preset stress parameters of the transistor to be simulated; acquiring equivalent pulse voltage according to degradation consistency conditions, the preset acceleration factor model, the Gaussian mixture interference signal time domain function, the gate oxide thickness of the transistor to be simulated, the effective starting moment of the periodic interference signal and the effective ending moment of the periodic interference signal; and determining the rectangular pulse signal according to the equivalent pulse voltage, the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
18. The reliability simulation device of claim 17, wherein the conversion module is further configured to: performing an accelerating electric field experiment on a plurality of test transistors at constant temperature, and obtaining the actual failure time of each test transistor by applying constant electric fields with different intensities to each test transistor, wherein the specification and the model of the plurality of test transistors are the same; taking the constant electric field of each test transistor and the corresponding actual failure time as a group of data, and fitting according to a plurality of groups of data to obtain a failure time model, wherein the specification and model of the test transistor are the same as those of the transistor to be simulated; acquiring preset stress parameters according to the failure time model and a group of data; and acquiring the preset acceleration factor model according to the preset stress parameter, the effective electric field static stress and the real electric field stress.
19. The reliability simulation device of claim 16 or 17, wherein the degradation consistency condition comprises:
wherein P is the preset time length,and t is time for the preset acceleration factor model, and the preset duration is determined according to the effective starting time of the periodic interference signal and the effective ending time of the periodic interference signal.
20. The reliability simulation device of claim 16, wherein the acquisition module is further configured to acquire a signal characteristic period of the electromagnetic interference signal; and acquiring a periodic interference signal from the electromagnetic interference signal according to the signal characteristic period.
21. The reliability simulation device of claim 16, wherein the simulation module is further configured to obtain a target gaussian mixture model corresponding to the number of peaks of the periodic interference signal according to the number of peaks; and performing Gaussian mixture simulation on the periodic interference signal according to the target Gaussian mixture model by adopting an expected maximum algorithm to obtain a Gaussian mixture interference signal time domain function.
22. The reliability simulation device of claim 21 wherein the simulation module is further configured to convert a voltage strength in a time domain of the periodic interference signal to a probability density in a time domain, the probability density being determined from a ratio of a voltage strength value to a total voltage strength value; obtaining optimal parameters of Gaussian mixture model probability functions corresponding to the target Gaussian mixture model according to preset distribution data, wherein the Gaussian mixture model probability functions are obtained by weighting and summing one or more Gaussian distribution functions, and the optimal parameters comprise weight, distribution mean value and standard deviation of each Gaussian distribution function; and generating the Gaussian mixture interference signal time domain function according to the Gaussian mixture model probability function and the optimal parameter.
23. The reliability simulation device of claim 22 wherein the simulation module is further configured to obtain different combinations of parameters including a weight, a distribution mean, and a standard deviation for each of the gaussian distribution functions; calculating the distribution probability of the preset distribution data belonging to each Gaussian distribution function under each parameter combination, and updating the parameter combinations according to the distribution probability; and determining a target parameter combination, wherein the target parameter combination is the parameter combination with the error value of the parameter combination before and after updating being smaller than a preset error value.
24. The reliability simulation device of claim 16, wherein the simulation module is further configured to determine a plurality of simulation points in the rectangular pulse signal; sequentially inputting voltages corresponding to the simulation points as gate voltages of the transistor to be simulated, and performing simulation for preset times, wherein the simulation is performed once when the input of the simulation points of the rectangular pulse signal is completed; and detecting the current drain current of the transistor to be simulated after the simulation for the preset times, and determining the degradation rate according to the current drain current and the initial drain current.
25. The reliability simulation device of claim 24 wherein the degradation rate is characterized by a first ratio of the present drain current and the initial drain current or wherein the degradation rate is characterized by a second ratio of a difference between the initial drain current and the present drain current to the initial drain current.
26. The reliability simulation apparatus according to claim 16, wherein the number of simulation points when simulation is performed with the rectangular pulse signal is smaller than the number of simulation points when simulation is performed with the periodic disturbance signal.
27. The reliability simulation device of claim 16, wherein the simulation software simulates according to a predetermined reliability simulation model, the reliability simulation model being determined according to the type of the transistor to be simulated.
28. The reliability simulation apparatus of claim 26 wherein the transistor to be simulated comprises a laterally diffused metal oxide semiconductor transistor and the reliability simulation model comprises a reactive diffusion model.
29. The reliability simulation device of claim 16 wherein a plurality of the electromagnetic pulse signals are generated by a plurality of gas insulated switchgear devices operating in parallel at the same frequency.
30. The reliability simulation device of claim 16, wherein the transistor to be simulated comprises a diode or a triode.
31. A simulation device comprising a memory and a processor, the memory having stored therein a computer program, characterized in that the computer program, when executed by the processor, causes the processor to perform the reliability simulation method of any of claims 1 to 15.
32. A non-transitory computer readable storage medium containing a computer program which, when executed by a processor, causes the processor to perform the reliability simulation method of any one of claims 1-15.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311550020.1A CN117852462A (en) | 2023-11-20 | 2023-11-20 | Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311550020.1A CN117852462A (en) | 2023-11-20 | 2023-11-20 | Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117852462A true CN117852462A (en) | 2024-04-09 |
Family
ID=90543966
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311550020.1A Pending CN117852462A (en) | 2023-11-20 | 2023-11-20 | Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117852462A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113468792A (en) * | 2021-07-22 | 2021-10-01 | 国网宁夏电力有限公司电力科学研究院 | Parameter correction method and device of electromagnetic transient simulation model and electronic equipment |
CN116010226A (en) * | 2022-12-23 | 2023-04-25 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Software system reliability simulation evaluation method and device and computer equipment |
CN116150967A (en) * | 2022-12-20 | 2023-05-23 | 中国船舶集团有限公司综合技术经济研究院 | Radar reliability modeling method under suppression interference |
CN116205065A (en) * | 2023-02-27 | 2023-06-02 | 温州大学乐清工业研究院 | Method for evaluating storage time-varying reliability of electromagnetic relay |
US20230306159A1 (en) * | 2020-12-03 | 2023-09-28 | Huawei Technologies Co., Ltd. | Simulation test method, apparatus, and system |
-
2023
- 2023-11-20 CN CN202311550020.1A patent/CN117852462A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20230306159A1 (en) * | 2020-12-03 | 2023-09-28 | Huawei Technologies Co., Ltd. | Simulation test method, apparatus, and system |
CN113468792A (en) * | 2021-07-22 | 2021-10-01 | 国网宁夏电力有限公司电力科学研究院 | Parameter correction method and device of electromagnetic transient simulation model and electronic equipment |
CN116150967A (en) * | 2022-12-20 | 2023-05-23 | 中国船舶集团有限公司综合技术经济研究院 | Radar reliability modeling method under suppression interference |
CN116010226A (en) * | 2022-12-23 | 2023-04-25 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Software system reliability simulation evaluation method and device and computer equipment |
CN116205065A (en) * | 2023-02-27 | 2023-06-02 | 温州大学乐清工业研究院 | Method for evaluating storage time-varying reliability of electromagnetic relay |
Non-Patent Citations (1)
Title |
---|
葛承垄;朱元昌;邸彦强;孟宪国;: "面向一类混合退化装备RUL预测的平行仿真技术", 北京理工大学学报, no. 04, 15 April 2019 (2019-04-15) * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102262206B (en) | Method for predicting negative bias temperature instability (NBTI) service life of pMOSFET (P-channel Metal Oxide Semiconductor Field Effect Transistor) device | |
Xiong et al. | A data-driven multi-scale extended Kalman filtering based parameter and state estimation approach of lithium-ion polymer battery in electric vehicles | |
US20120065920A1 (en) | Evaluation method, evaluation apparatus, and simulation method of semiconductor device | |
CN103105571B (en) | Simulated measurement method of current characteristics of insulated gate bipolar transistor | |
CN106570293B (en) | A kind of method for calculating probability of the circuit signal reliability based on EPTM model | |
US10527665B2 (en) | Predicting noise propagation in circuits | |
JPH0799302A (en) | Simulation method of deterioration of hot carrier in mos type transistor | |
CN104316855A (en) | Junction temperature testing method of HEMT (High Electron Mobility Transistor) device | |
JP4500338B2 (en) | Test pattern evaluation method and test pattern evaluation apparatus | |
Ye et al. | A new class of multi-stress acceleration models with interaction effects and its extension to accelerated degradation modelling | |
CN101303390B (en) | Method for judging MOS device performance degeneration | |
US10628541B2 (en) | Method and apparatus for selecting integrated circuit device neural network modeling sample | |
CN109033534A (en) | Follower timing jitter estimation method based on pseudo- open-drain termination | |
US9519015B2 (en) | Rise time and fall time measurement | |
CN117852462A (en) | Reliability simulation method, reliability simulation device, simulation apparatus, and storage medium | |
US8271254B2 (en) | Simulation model of BT instability of transistor | |
CN108233356B (en) | Consistency evaluation method and evaluation platform for photovoltaic inverter controller | |
CN104849647A (en) | Method and device for simulating reliability aging of circuit | |
CN116150594A (en) | Method for identifying switch element characteristics in spectrum test data | |
CN107576866A (en) | It is a kind of to approach sparse reconstruct and the method for distinguishing multiple harmonic sources of interior point method based on smooth | |
Pichler et al. | Experimental study on load dependent X-parameter models for PA design | |
CN110658436A (en) | Characterization method for MOS transistor performance degradation under radio frequency stress | |
Dutta et al. | BSIM6--Benchmarking the Next-Generation MOSFET Model for RF Applications | |
Leuchter et al. | Automatic test-bench for SiC power devices using LabVIEW | |
CN108614904A (en) | Parametric Yield of VLSI IC rapid analysis method based on emulation data dependence |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |