A highly integrated micro intelligent fuse module and its manufacturing process

Abstract

The present invention discloses a highly integrated micro intelligent fuse module and its manufacturing process, including performing high-frequency electromagnetic disturbance detection on the interface area between the fusing element and the temperature sensor to obtain the electromagnetic response fingerprint and determine the suspected interface defect point distribution map; applying a micro-pulse thermal excitation to the fusing element, capturing the thermal pulse propagation process and the sensor response, and generating the interface thermal resistance distribution map; performing acoustic microscopy analysis on the area with abnormal thermal resistance to generate a fine classification map of interface defects; performing a collaborative test on the electro-thermal transient response of the fuse module to obtain the performance parameter test results; applying targeted accelerated aging treatment to different types of interface defect samples, and periodically performing detections during the aging process to determine the quality grade and reliability prediction results. The present invention realizes the identification and characterization of defects in the interface area of the intelligent fuse module, establishes the correlation between defect characteristics and functional impacts, effectively evaluates the long-term reliability of the product, and significantly improves the product quality control level.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic component manufacturing and detection, and in particular to a highly integrated miniature intelligent fuse module and a preparation process thereof. Background Art

[0002] The highly integrated miniature intelligent fuse module is a new type of circuit protection device. Its structure primarily consists of a micro-fuse element, a temperature sensor array, an intelligent control chip, and a communication interface. This module integrates a traditional fuse link with an intelligent control system within a package (typically measuring no more than 5×5×2mm). Its operating principle is to use a distributed temperature sensor array to monitor the temperature distribution of the fuse element in real time. The intelligent control chip analyzes the current waveform characteristics and temperature change patterns, proactively triggering the protection mechanism before a dangerous condition occurs. The fuse element is precision-cold-rolled and stretched to a 100μm diameter using a special alloy. It connects to the temperature sensor via a microspring for signal transmission. The entire system is encapsulated in epoxy resin with nano-thermal conductive fillers to improve heat dissipation.

[0003] During the manufacturing process of smart fuse modules, the epoxy resin potting step significantly impacts product quality. Due to the module's complex internal structure and tiny size, the epoxy resin, which is added with nano-thermally conductive fillers, has reduced fluidity during the potting process, easily forming microbubbles and uneven filler distribution at the interface between the fuse element and the temperature sensor. These microscopic defects in the interface significantly reduce the local thermal conductivity, preventing the temperature sensor from accurately capturing the actual temperature state of the fuse element, leading to protection delays or malfunctions. Furthermore, the volume shrinkage effect of the epoxy resin during curing creates stress concentration zones around the microbubbles. This, coupled with temperature cycling, promotes the formation and propagation of microcracks, ultimately causing the fuse module's protection to fail. Traditional inspection methods such as X-rays and ultrasonic scanning have limited resolution for micron-level defect detection, making it difficult to effectively identify these interface microdefects that significantly impact product functionality, especially when these defects are located in critical heat conduction paths. Therefore, developing a high-precision nondestructive testing method to identify and characterize microdefects and their thermal response characteristics at the interface between the fuse element and the temperature sensor in fuse modules is a key technical issue for ensuring product functional integrity and reliability. Summary of the Invention

[0004] The main purpose of the present invention is to solve the technical problems of insufficient resolution, inability to characterize thermal response characteristics and lack of function-oriented evaluation in the existing interface micro-defect detection method during the preparation process of smart fuse modules.

[0005] A first aspect of the present invention provides a process for preparing a highly integrated miniature intelligent fuse module. The process for preparing the highly integrated miniature intelligent fuse module comprises:

[0006] Perform high-frequency electromagnetic disturbance detection on the interface area between the fusing element and the temperature sensor of the fuse module, obtain the electromagnetic response fingerprint of the interface area, and determine the suspected interface defect point distribution map based on the electromagnetic response fingerprint;

[0007] According to the suspected interface defect point distribution map, apply a micro-pulse thermal excitation to the fusing element, capture the thermal pulse propagation process and the temperature sensor response, and generate an interface thermal resistance distribution map;

[0008] According to the thermal resistance abnormal area in the interface thermal resistance distribution map, perform acoustic microscopy analysis, collect multi-dimensional acoustic parameters, and generate a fine classification map of interface defects;

[0009] According to the fine classification map of interface defects, conduct an electro-thermal transient response collaborative test on the fuse module, record the temperature rise characteristics of the fusing element, the interface temperature change, and the sensor response under various current conditions, and obtain the performance parameter test results;

[0010] According to the performance parameter test results, apply targeted accelerated aging treatment to samples with different types of interface defects, perform periodic detection during the aging process, track the defect evolution process and the performance degradation trend, and determine the quality grade and reliability prediction results.

[0011] Preferably, the performing high-frequency electromagnetic disturbance detection on the interface area between the fusing element and the temperature sensor of the fuse module, obtaining the electromagnetic response fingerprint of the interface area, and determining the suspected interface defect point distribution map includes:

[0012] Apply an electromagnetic disturbance with a first frequency in the range of 2.0 - 3.0 GHz to the fusing element area of the fuse module, and apply an electromagnetic disturbance with a second frequency in the range of 5.0 - 6.0 GHz to the temperature sensor area;

[0013] Capture the electromagnetic reflection wave and transmission wave of the interface area between the fusing element and the temperature sensor, record the reflection coefficient and phase change of the electromagnetic reflection wave, and the transmission coefficient and phase delay of the transmission wave;

[0014] Conduct multi-parameter fusion analysis on the reflection coefficient, phase change, transmission coefficient, and phase delay to generate the electromagnetic response fingerprint of the interface area;

[0015] Compare the electromagnetic response fingerprint with the standard reference value, and calculate the anomaly index according to the deviation degree;

[0016] Identify the characteristic patterns corresponding to micro-bubbles, uneven filler distribution, and micro-cracks according to the anomaly index and the characteristic parameters in the electromagnetic response fingerprint, and determine the suspected interface defect point distribution map.

[0017] Preferably, the multi-parameter fusion analysis of the reflection coefficient, phase change, transmission coefficient and phase delay to generate the electromagnetic response fingerprint of the interface region includes:

[0018] Divide the interface region between the fusing element and the temperature sensor into three functional structural regions: the metal-epoxy resin transition region, the epoxy resin main body region, and the semiconductor-epoxy resin transition region, and extract the reflection coefficient, phase change, transmission coefficient and phase delay for each functional structural region respectively;

[0019] Based on the distribution characteristics of nano-thermal conductive fillers in epoxy resin, conduct a composite analysis of the reflection coefficient and phase change in the metal-epoxy resin transition region to generate metal interface characteristic parameters;

[0020] Based on the dielectric properties of epoxy resin, conduct a correlation analysis of the transmission coefficient and phase delay in the epoxy resin main body region to generate filler distribution characteristic parameters;

[0021] Based on the interface characteristics between semiconductor materials and epoxy resin, conduct a comparative analysis of the reflection coefficient and transmission coefficient in the semiconductor-epoxy resin transition region to generate sensor interface characteristic parameters;

[0022] Integrate and map the metal interface characteristic parameters, filler distribution characteristic parameters and sensor interface characteristic parameters to generate the electromagnetic response fingerprint of the interface region.

[0023] Preferably, the identification of the characteristic patterns corresponding to microbubbles, uneven filler distribution and microcracks and the determination of the suspected interface defect point distribution map according to the abnormal index and the characteristic parameters in the electromagnetic response fingerprint include:

[0024] Conduct a fusing heat flow transfer path analysis on the abnormal index, and divide the interface region into a key heat transfer region, a secondary heat transfer region and a non-heat transfer region, where the region where the fusing element is in direct contact with the temperature sensor and the distance is less than 100 microns is defined as the key heat transfer region, the region with a distance of 100 - 300 microns from the edge of the key heat transfer region is defined as the secondary heat transfer region, and the region with a distance greater than 300 microns from the edge of the key heat transfer region is defined as the non-heat transfer region;

[0025] Calculate the covariance matrix of the phase delay and reflection coefficient in each region of the electromagnetic response fingerprint, extract the eigenvalue distribution pattern from the covariance matrix, and determine the position of microbubbles by identifying the double-peak feature in the eigenvalue distribution pattern;

[0026] Calculate the spatial gradient of the transmission coefficient in the electromagnetic response fingerprint, analyze the anisotropy difference of the spatial gradient, and determine the uneven filler distribution region according to the region where the value of the anisotropy difference is greater than the preset threshold;

[0027] Perform a differential operation on the electromagnetic responses of the first frequency and the second frequency to obtain a difference signal, extract the linear discontinuity features in the difference signal, and determine the positions and orientations of microcracks based on the linear discontinuity features;

[0028] Calculate the hazard level value based on the distribution of the microbubble positions, the regions with uneven filler distribution, the positions and orientations of the microcracks in the critical heat transfer region, the secondary heat transfer region, and the non - heat transfer region, and in combination with the degree of blocking of the heat flow path. Map the microbubble positions, the regions with uneven filler distribution, the positions and orientations of the microcracks, and the hazard level value to the interface region coordinate system to generate a suspected interface defect point distribution map.

[0029] Preferably, according to the suspected interface defect point distribution map, applying a micro - pulse thermal excitation to the fuse element, capturing the heat pulse propagation process and the temperature sensor response, and generating an interface thermal resistance distribution map, including:

[0030] According to the suspected interface defect point distribution map, determine the key path through which the heat flow passes, set up a thermal excitation point array on the key path, increase the density of thermal excitation points in the region of the thermal excitation point array close to the microbubble position, adjust the power parameters of the thermal excitation points close to the region with uneven filler distribution, and extend the pulse duration of the thermal excitation points passing through the microcrack region;

[0031] Apply micro - pulse thermal excitation point - by - point to the thermal excitation point array corresponding to the fuse element;

[0032] Collect the spatio - temporal evolution data of the temperature field for each power thermal pulse of each thermal excitation point, record the temperature distribution, propagation speed, and attenuation characteristics of the heat pulse propagating from the fuse element to the temperature sensor, and generate a temperature field evolution data set;

[0033] Synchronously collect the electrical signal response curves generated by the temperature sensor for each power thermal pulse of each thermal excitation point, calculate the temperature - electrical signal transfer function based on the temperature field evolution data set and the electrical signal response curves, and obtain the sensor response characteristic parameters;

[0034] Perform Fourier thermal wave analysis on the temperature field evolution data set, and calculate the thermal resistance value and thermal diffusivity of the region corresponding to each thermal excitation point based on the attenuation of the thermal wave amplitude and the phase delay;

[0035] Map the thermal resistance value, thermal diffusivity, and the sensor response characteristic parameters to the spatial coordinate system of the interface region, perform a spatial correlation analysis on the thermal resistance value, thermal diffusivity, and the suspected interface defect point distribution map, and generate an interface thermal resistance distribution map including thermal conduction performance and sensor response performance.

[0036] Preferably, based on the abnormal thermal resistance regions in the interface thermal resistance distribution map, acoustic microscopy analysis is performed to collect multi-dimensional acoustic parameters and generate a fine classification map of interface defects, including:

[0037] Based on the interface thermal resistance distribution map, identify the regions where the thermal resistance value exceeds 30% of the reference value as the key regions for acoustic detection, perform acoustic scanning on the key regions for acoustic detection, and obtain four basic acoustic parameters: acoustic wave reflection intensity, acoustic wave transmission intensity, sound velocity change, and acoustic wave phase delay;

[0038] Perform correlation analysis on the basic acoustic parameters and the material properties of the interface region. Analyze the acoustic wave reflection intensity and sound velocity change for the metal-epoxy resin interface region, analyze the acoustic wave transmission intensity and phase delay for the epoxy resin filler region, and analyze the acoustic wave reflection mode and phase jump characteristics for the temperature sensor-epoxy resin interface region to generate defect acoustic feature data;

[0039] Perform pattern recognition on the defect acoustic feature data. Based on the strong reflection-weak transmission combination feature of microbubbles, the medium reflection-nonlinear phase delay combination feature of uneven filler distribution, and the directional reflection-sharp phase jump combination feature of microcracks, determine the type, location, size, and severity of the defects, and generate a fine classification map of interface defects.

[0040] Preferably, based on the fine classification map of interface defects, perform electro-thermal transient response co-testing on the fuse module, record the temperature rise characteristics of the fuse element, interface temperature changes, and sensor responses under various current conditions, and obtain the test results of performance parameters, including:

[0041] According to the defect types and distribution positions in the fine classification map of interface defects, divide the fuse module into a defect-free region, a region affected by microbubbles, a region affected by uneven filler distribution, and a region affected by microcracks;

[0042] Apply a standard current condition to the defect-free region to obtain reference performance data, apply a current slow-rise condition to the region affected by microbubbles, apply a current step condition to the region affected by uneven filler distribution, apply a current pulse condition to the region affected by microcracks, and record the temperature rise characteristic curves of the fuse elements in each region;

[0043] Collect the interface temperature change data of the defect-free region, the region affected by microbubbles, the region affected by uneven filler distribution, and the region affected by microcracks under their respective current conditions, calculate the temperature gradient and heat flux distribution of each region, and obtain a dynamic distribution map of interface temperature;

[0044] Synchronously record the temperature sensor response signals of the defect-free area, the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks under their respective current conditions, calculate the temperature transfer efficiency parameter in combination with the temperature rise characteristic curve of the fuse element, and calculate the response time parameter and the temperature-signal distortion parameter in combination with the dynamic interface temperature distribution map;

[0045] Compare and analyze the reference performance data with the performance data of the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks, establish the correlation between the defect type and the temperature transfer efficiency parameter, the response time parameter, and the temperature-signal distortion parameter, and obtain the test results of the performance parameters.

[0046] Preferably, according to the test results of the performance parameters, apply targeted accelerated aging treatment to the samples with different types of interface defects, perform periodic detection during the aging process, track the defect evolution process and the performance degradation trend, and determine the quality grade and the reliability prediction result, including:

[0047] According to the degree of influence of the defect type on the performance in the test results of the performance parameters, apply temperature cycle aging to the samples affected by microbubbles, apply high temperature and high humidity aging to the samples affected by uneven filler distribution, and apply temperature-current composite cycle aging to the samples affected by microcracks to generate an aging sample set;

[0048] Perform interface thermal resistance imaging, acoustic microscopy analysis, and electro-thermal transient response testing on the key nodes of the aging sample set during the aging process to obtain defect size change data, interface structure change data, and performance parameter change data;

[0049] Perform correlation analysis on the defect size change data, interface structure change data, and performance parameter change data, and calculate the defect size growth rate and the hazard coefficient of various defects;

[0050] According to the defect size growth rate and the hazard coefficient of various defects, classify the micro intelligent fuse module into four quality grades: A, B, C, and D, calculate the expected life of each grade of samples, and determine the quality grade and the reliability prediction result.

[0051] The second aspect of the present invention provides a highly integrated micro intelligent fuse module, and the preparation process of the highly integrated micro intelligent fuse module adopts the above-mentioned preparation process of the highly integrated micro intelligent fuse module.

[0052] The technical solution provided by the embodiments of this application obtains the electromagnetic response fingerprint of the interface area through high-frequency electromagnetic perturbation detection, and identifies the suspected interface defect points based on the unique response characteristics of different materials to electromagnetic waves. Subsequently, a precisely controlled micro-pulse thermal excitation is applied to the fusing element to simulate the heat generation under actual working conditions, while capturing the heat conduction process and the sensor response to generate a thermal resistance distribution map. Then, acoustic microscopy analysis is performed on the area with abnormal thermal resistance, and the microscopic structure of the defect is characterized by using the reflection and transmission characteristics of sound waves at the interfaces of different media. Then, an electro-thermal transient response collaborative test is carried out to directly evaluate the actual impact of the defect on the function of the fuse. Finally, targeted accelerated aging treatment is carried out on different types of defect samples to track the defect evolution and performance degradation trends, and determine the quality grade and reliability prediction results.

[0053] This series of steps constitutes a complete defect detection and evaluation chain, from the discovery, characterization, functional impact evaluation of interface micro-defects to long-term reliability prediction, forming a closed-loop solution. High-frequency electromagnetic perturbation detection realizes the non-destructive discovery of micron-level defects, overcoming the limitation of the resolution of traditional X-ray detection; the micro-pulse thermal excitation not only characterizes the physical properties of the defect, but also directly reveals its impact on the heat conduction function, solving the problem that static detection cannot reflect dynamic functions; acoustic microscopy analysis provides accurate information on the type and morphology of the defect, providing a basis for targeted treatment; the electro-thermal transient response collaborative test directly correlates the defect with the function, overcoming the limitation of traditional detection that only focuses on the defect itself and ignores the functional impact; accelerated aging and periodic detection solve the problem of predicting the long-term evolution of defects. This multi-dimensional, multi-scale, function-oriented detection method effectively overcomes the deficiencies of traditional detection in the identification and evaluation of interface area defects in micro intelligent fuse modules. Brief Description of the Drawings

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0055] Figure 1 It is a schematic diagram of an embodiment of the preparation process of a highly integrated micro intelligent fuse module in the embodiments of the present invention.

[0056] The realization of the objectives, functional characteristics and advantages of the present invention will be further described with reference to the embodiments and the drawings. Detailed Embodiments

[0057] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0058] It should be noted that if there are directional indications (such as up, down, left, right, front, back,...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative position relationship, movement conditions, etc. between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly.

[0059] In addition, the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, "and / or" throughout the text includes three scenarios. Taking A and / or B as an example, it includes the technical solution of A, the technical solution of B, and the technical solution that both A and B are satisfied at the same time. In addition, the technical solutions between various embodiments can be combined with each other, which must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0060] An embodiment of the present application provides a preparation process for a highly integrated micro intelligent fuse module. Figure 1 It is a flowchart of a preparation process for a highly integrated micro intelligent fuse module provided by an embodiment of the present application. In this embodiment, the method includes:

[0061] Please refer to Figure 1 , perform high-frequency electromagnetic disturbance detection on the interface area between the fuse element and the temperature sensor of the fuse module, obtain the electromagnetic response fingerprint of the interface area, and determine the distribution map of suspected interface defect points based on the electromagnetic response fingerprint;

[0062] In an embodiment of the present invention, the performing high-frequency electromagnetic disturbance detection on the interface area between the fuse element and the temperature sensor of the fuse module, obtaining the electromagnetic response fingerprint of the interface area, and determining the distribution map of suspected interface defect points based on the electromagnetic response fingerprint includes:

[0063] Apply an electromagnetic disturbance with a first frequency in the range of 2.0 - 3.0 GHz to the fuse element area of the fuse module, and apply an electromagnetic disturbance with a second frequency in the range of 5.0 - 6.0 GHz to the temperature sensor area;

[0064] capturing electromagnetic reflected waves and transmitted waves in the interface area between the fuse element and the temperature sensor, and recording the reflection coefficient and phase change of the electromagnetic reflected waves, and the transmission coefficient and phase delay of the transmitted waves;

[0065] performing a multi-parameter fusion analysis on the reflection coefficient, phase change, transmission coefficient and phase delay to generate an electromagnetic response fingerprint of the interface region;

[0066] Comparing the electromagnetic response fingerprint with a standard reference value and calculating an anomaly index based on the degree of deviation;

[0067] According to the abnormality index and the characteristic parameters in the electromagnetic response fingerprint, characteristic patterns corresponding to microbubbles, uneven filler distribution and microcracks are identified, and a distribution map of suspected interface defect points is determined.

[0068] Specifically, a micro-directional antenna array is used to apply a first frequency electromagnetic perturbation in the 2.0-3.0 GHz range to the fuse element area of the fuse module, and a second frequency electromagnetic perturbation in the 5.0-6.0 GHz range to the temperature sensor area. This step is based on the differences in the response of different materials to electromagnetic waves, and the frequency parameters are selected in a targeted manner: the 2.0-3.0 GHz frequency band has a moderate penetration depth into the metal fuse element, which can effectively identify defects inside and on the surface of the metal; while the 5.0-6.0 GHz frequency band is highly sensitive to semiconductor materials, which can better detect anomalies in the temperature sensor area.

[0069] Capturing the electromagnetic reflected and transmitted waves at the interface between the fuse element and the temperature sensor, and recording the relevant parameters, is accomplished using a high-sensitivity receiving array. The reflection coefficient represents the proportion of the incident wave reflected by the interface, measured in decibels (dB); the phase change refers to the change in phase of the electromagnetic wave during the reflection process, expressed in degrees; the transmission coefficient represents the proportion of energy that passes through the interface; and the phase delay reflects the degree of phase delay of the wave after passing through the interface. For example, when microbubbles are present in the interface area, the reflection coefficient will increase significantly (typically by 6-10dB), and the phase change will jump (typically 150°-180°); and when there is an uneven distribution of fillers, the transmission coefficient will show irregular attenuation (usually within the range of ±4dB).

[0070] Multi-parameter fusion analysis of the reflection coefficient, phase change, transmission coefficient, and phase delay is to map these four parameters into a unified feature space and generate an electromagnetic response fingerprint through a weighted combination algorithm. In the specific process, first, each parameter is normalized to eliminate the dimensional difference; then, weights are assigned according to the sensitivity of different parameters to various types of defects. For example, for microbubble detection, the weight of the reflection coefficient is 0.4, the phase change is 0.3, the transmission coefficient is 0.2, and the phase delay is 0.1; finally, sensitive features are extracted through a non-linear transformation to generate a multi-dimensional electromagnetic feature matrix. For example, in a test case, it is found that the comprehensive fingerprint value in the central region of the interface is 35% higher than that in the surrounding regions, indicating an abnormality in this region.

[0071] Comparing the electromagnetic response fingerprint with the standard reference value and calculating the anomaly index is achieved by comparing the measured fingerprint with the fingerprint library of defect-free samples established in advance. For each spatial point, the Euclidean distance between the measured value and the reference value is calculated, and the anomaly index is defined according to the degree of deviation. The anomaly index is usually a value between 0 and 1, where 0 indicates a perfect match with the reference value and 1 indicates the maximum deviation. In practice, regions with an anomaly index greater than 0.3 are usually marked as regions of concern, and regions greater than 0.6 are marked as high-risk regions.

[0072] Identifying the defect feature pattern based on the anomaly index and the characteristic parameters of the electromagnetic response fingerprint is achieved through a pattern recognition algorithm. Different types of defects present unique patterns in the electromagnetic response: microbubbles appear as circular or elliptical regions with local strong reflection and phase mutation; uneven filler distribution shows a gradual anomaly in the transmission coefficient and a non-linear change in the phase delay; microcracks show linear reflection enhancement and directional phase perturbation. For example, in an actual case, a linear anomaly region about 200 μm long is observed, with a reflection enhancement of about 8 dB and the phase perturbation extending along a specific direction, which is accurately identified as an interface microcrack.

[0073] In an embodiment of the present invention, the multi-parameter fusion analysis of the reflection coefficient, phase change, transmission coefficient, and phase delay to generate the electromagnetic response fingerprint of the interface region includes:

[0074] The interface region between the fuse element and the temperature sensor is divided into three functional structural regions: the metal-epoxy resin transition region, the epoxy resin main body region, and the semiconductor-epoxy resin transition region. The reflection coefficient, phase change, transmission coefficient, and phase delay are respectively extracted for each functional structural region;

[0075] Based on the distribution characteristics of the nano-thermal conductive filler in the epoxy resin, a composite analysis of the reflection coefficient and phase change in the metal-epoxy resin transition region is performed to generate metal interface characteristic parameters;

[0076] Based on the dielectric properties of the epoxy resin, a correlation analysis is performed on the transmission coefficient and phase delay of the epoxy resin main body region to generate filler distribution characteristic parameters;

[0077] Based on the interface properties between the semiconductor material and the epoxy resin, a comparative analysis is performed on the reflection coefficient and transmission coefficient of the semiconductor-epoxy resin transition region to generate sensor interface characteristic parameters;

[0078] Integrate and map the metal interface characteristic parameters, filler distribution characteristic parameters, and sensor interface characteristic parameters to generate the electromagnetic response fingerprint of the interface region.

[0079] Specifically, dividing the interface region of the fuse element and the temperature sensor into three functional structural regions is achieved by using a spatial positioning scanning system. The specific method is as follows: First, determine the geometric center point of the interface based on the structural design drawing of the fuse module; then, taking this as a reference, a range of 100 microns extending towards the fuse element is defined as the metal-epoxy resin transition region, and a range of 100 microns extending towards the temperature sensor is defined as the semiconductor-epoxy resin transition region, and the region between the two transition regions is defined as the epoxy resin main body region. High-frequency electromagnetic scanning is performed on each region separately to extract electromagnetic parameters. This region division method takes into account the special properties of the interfaces between different materials. There are obvious differences in the interface properties between metal and epoxy resin, and semiconductor and epoxy resin. Zonal detection can adopt the best parameter combination according to different interface characteristics.

[0080] The composite analysis of the reflection coefficient and phase change of the metal-epoxy resin transition region is achieved through a correlation analysis algorithm. The metal-epoxy resin transition region is the contact area between the fuse element and the filled epoxy resin. The distribution of nano-thermal conductive fillers in this region directly affects the heat transfer from the fuse element to the epoxy resin. The specific operation is as follows: First, normalize the reflection coefficient data and phase change data of this region; then calculate the spatial correlation and frequency correlation between the two; finally, apply a convolution algorithm to generate metal interface characteristic parameters. In actual tests, when there are microcracks at the metal-epoxy resin interface, the correlation coefficient between the reflection coefficient and the phase change will drop from the normal 0.85 to below 0.4; when there is filler aggregation, it is manifested as an increase in the reflection coefficient and local fluctuations in the phase change. The metal interface characteristic parameters are crucial for evaluating the integrity of the fuse element interface because this interface is the starting point of heat generation and conduction, and its quality directly determines the thermal response characteristics of the fuse.

[0081] The correlation analysis of the transmission coefficient and phase delay in the epoxy resin main region is achieved through multi-dimensional spectrum analysis. The dielectric properties of epoxy resin (the dielectric constant is usually between 3.2 - 4.5) directly affect the propagation characteristics of electromagnetic waves, and adding nano-thermal conductive fillers will change this property. The specific method is as follows: perform Fourier transform on the transmission coefficient and phase delay data to analyze their spectral characteristics; calculate the phase difference between the two at different spatial positions and frequency points; finally, generate the filler distribution characteristic parameters through weighted fusion. For example, in a test sample, the area with uniform nano-filler distribution in the epoxy resin region shows a smooth transmission coefficient curve (change less than ±1 dB) and a linear phase delay curve; while the area with filler aggregation or absence shows a local mutation of the transmission coefficient (change greater than ±3 dB) and a non-linear change of the phase delay. This analysis can effectively identify abnormal filler distribution in epoxy resin, which has a direct impact on the heat dissipation performance and heat conduction uniformity of the fuse.

[0082] The comparative analysis of the reflection coefficient and transmission coefficient in the semiconductor-epoxy resin transition region is achieved through the differential comparison algorithm. The interface between semiconductor materials (such as silicon or gallium arsenide) and epoxy resin shows unique characteristics in the response to electromagnetic waves. The specific operation is as follows: calculate the ratio of the reflection coefficient to the transmission coefficient at each measurement point in this region; analyze the spatial distribution pattern of this ratio; construct a reflection-transmission ratio map and compare it with the standard template to generate the sensor interface characteristic parameters. In actual detection, a good semiconductor-epoxy resin interface shows a stable distribution of the reflection / transmission ratio within the range of 0.8 - 1.2; when there are microbubbles at the interface, this ratio will suddenly increase to more than 2.0 at the bubble position; poor interface bonding is manifested as large-area ratio anomalies. The quality of the semiconductor-epoxy resin interface directly affects the response accuracy of the temperature sensor, which is a key link in the realization of the functions of the micro intelligent fuse.

[0083] The integration mapping of the characteristic parameters of the three functional regions is achieved through the spatial feature fusion system. This system first establishes a unified three-dimensional coordinate system for the interface region; then maps the metal interface characteristic parameters, filler distribution characteristic parameters, and sensor interface characteristic parameters to this coordinate system according to their respective spatial positions; finally, through the spatial interpolation algorithm, generate the complete electromagnetic response fingerprint of the interface region. An electromagnetic response fingerprint is a three-dimensional data set, containing position information on the x-y plane and the comprehensive characteristic value on the z-axis. For example, in the detected electromagnetic response fingerprint, the characteristic values in the normal region fluctuate no more than ±10% of the reference value, while the defective region shows significant peaks (microbubbles) or valleys (filler absence). This integration mapping method comprehensively considers the characteristics of different functional regions, forms a comprehensive and accurate characterization of the entire interface region, and provides a multi-dimensional information basis for subsequent defect identification.

[0084] In an embodiment of the present invention, identifying the characteristic patterns corresponding to microbubbles, uneven filler distribution, and microcracks according to the characteristic parameters in the abnormal index and the electromagnetic response fingerprint, and determining the suspected interface defect point distribution map includes:

[0085] Conduct a fusing heat flow transfer path analysis on the abnormal index, and divide the interface area into a key heat transfer area, a secondary heat transfer area, and a non-heat transfer area. Among them, the area where the fusing element is in direct contact with the temperature sensor and the distance is less than 100 microns is defined as the key heat transfer area, the area with a distance of 100 - 300 microns from the edge of the key heat transfer area is defined as the secondary heat transfer area, and the area with a distance greater than 300 microns from the edge of the key heat transfer area is defined as the non-heat transfer area;

[0086] Calculate the covariance matrix for the phase delay and reflection coefficient of each area in the electromagnetic response fingerprint, extract the eigenvalue distribution pattern from the covariance matrix, and determine the position of microbubbles by identifying the double-peak feature in the eigenvalue distribution pattern;

[0087] Perform a spatial gradient calculation on the transmission coefficient in the electromagnetic response fingerprint, analyze the anisotropy difference of the spatial gradient, and determine the uneven filler distribution area according to the area where the value of the anisotropy difference is greater than the preset threshold;

[0088] Perform a differential operation on the electromagnetic responses at the first frequency and the second frequency to obtain a difference signal, extract the linear discontinuity feature in the difference signal, and determine the position and orientation of the microcrack according to the linear discontinuity feature;

[0089] According to the distribution of the microbubble position, the uneven filler distribution area, the microcrack position and orientation in the key heat transfer area, the secondary heat transfer area, and the non-heat transfer area, calculate the hazard level value in combination with the degree of blocking of the heat flow path, and map the microbubble position, the uneven filler distribution area, the microcrack position and orientation, and the hazard level value to the interface area coordinate system to generate a suspected interface defect point distribution map.

[0090] Specifically, the analysis of the heat transfer path for fusing due to the abnormal index is achieved through a heat conduction simulation system. The specific operations are as follows: First, import the structural design parameters of the fuse module and the data of the material's heat conduction characteristics. Then, based on the abnormal index distribution, simulate the heat transfer path after the fusing element generates heat. Finally, according to the spatial distribution of the heat flux density, divide the interface region into three regions. The key heat transfer region refers to the region where the fusing element is in direct contact with the temperature sensor and the distance is less than 100 micrometers. This region undertakes more than 80% of the heat transfer tasks. The secondary heat transfer region refers to the region 100 - 300 micrometers away from the edge of the key heat transfer region, which undertakes about 15 - 20% of the heat transfer function. The non - heat transfer region refers to the region more than 300 micrometers away from the edge of the key heat transfer region, where the heat flux density is extremely low. For example, in a 5×5 mm micro - intelligent fuse module, the area of the key heat transfer region is about 0.5×0.5 mm, and the secondary heat transfer region surrounds the key region in a ring shape with a width of about 200 micrometers. This method of region division based on the heat transfer path takes into account the actual working mechanism of the fuse. There are significant differences in the degree of influence of defects in different regions on the fuse function, and the region division makes the subsequent defect assessment more targeted.

[0091] Calculating the covariance matrix for the phase delay and reflection coefficient in each region of the electromagnetic response fingerprint is completed through statistical analysis software. The specific method is as follows: Normalize the phase delay and reflection coefficient data at each spatial position. Calculate the covariance matrix (a 2×2 matrix) of the two. Perform eigenvalue decomposition on this matrix to obtain two eigenvalues. Analyze the spatial distribution pattern of the eigenvalues. When there are micro - bubbles in the interface region, the gas - solid two - phase interface will produce an obvious bimodal feature in the eigenvalue distribution - the ratio of the two eigenvalues is greater than 3.0, and the spatial positions are concentrated. For example, in a sample containing micro - bubbles with a diameter of 20 micrometers, the eigenvalue ratio at the bubble position is 4.2, which is much higher than 1.2 - 1.8 in the surrounding regions. This method of covariance matrix analysis is particularly suitable for identifying micro - bubbles because the presence of bubbles will cause strong reflection and phase jumps of electromagnetic waves at the gas - solid interface, forming unique covariance characteristics. Through this method, micro - bubbles with a diameter as small as 10 micrometers can be identified.

[0092] The spatial gradient calculation of the transmission coefficient in the electromagnetic response fingerprint is achieved through an image processing system. The specific steps include: converting the transmission coefficient data into a two-dimensional grayscale image; applying the Sobel operator to calculate the gradients in the x and y directions; calculating the gradient magnitude and direction angle; analyzing the anisotropic differences of the gradients (the change ratio of the gradient magnitudes in different directions). The anisotropic difference refers to the degree of difference in the gradient values in different directions, and the larger the value, the stronger the non-uniformity of the material properties in the spatial distribution. In the region where the fillers are evenly distributed, the anisotropic difference is usually less than 0.3; while in the regions where the fillers are aggregated or missing, the anisotropic difference can reach 0.6 - 0.8. For example, in an epoxy resin sample with a nano-filler mass fraction of 5%, the anisotropic difference in the evenly filled region is 0.25, while in an obvious aggregation region, this value is 0.72. Setting the preset threshold to 0.5 is a critical value determined based on experimental data, which can effectively distinguish the normal region from the region with uneven filler distribution. This spatial-gradient-based analysis method is particularly suitable for detecting uneven filler distribution because the aggregation or absence of fillers will change the local electromagnetic properties, forming gradient variations.

[0093] The differential operation of the electromagnetic responses at the first frequency and the second frequency is achieved through a signal processing system. The specific operations include: spatially registering the electromagnetic response data obtained at the two frequencies; performing a pixel-level differential operation to obtain the difference signal; applying an edge detection algorithm to the difference signal to identify linear discontinuity features; using the Hough transform to extract the linear features and determine the position and orientation of the microcracks. Due to its linear structural characteristics, microcracks will produce frequency-selective responses under the action of electromagnetic waves at different frequencies, so they appear as linear discontinuity features in the differential signal. For example, in a sample containing microcracks with a length of 150 microns and a width of 3 microns, the differential signal at 2.4 GHz and 5.6 GHz clearly shows a linear anomaly, while this feature is not obvious in the single-frequency signals in the same region. This differential analysis method utilizes the response differences of electromagnetic waves at different frequencies to microcracks and can effectively identify tiny cracks that are difficult to detect by traditional single-frequency detection.

[0094] Calculating the hazard level value based on the distribution of various defects in different heat transfer regions is accomplished through a thermal-mechanical coupling analysis system. The specific steps include: establishing a correlation model between the defect type and the impact on heat conduction; assigning weights according to the importance of the defect location in the heat flow path (critical heat transfer region: 1.0, secondary heat transfer region: 0.6, non-heat transfer region: 0.2); considering the degree of influence of the defect size and shape on heat flow blockage; and comprehensively calculating the hazard level value (0 - 100). For example, when a microbubble with a diameter of 30 microns is located in the critical heat transfer region, the hazard level value is 75; when a bubble of the same size is located in the non-heat transfer region, the hazard level value is only 15. Mapping the defect information and the hazard level value to the interface region coordinate system is achieved through a visualization system. The generated interface defect suspected point distribution map uses pseudo-color coding, with different colors representing different types of defects and hazard levels. This analysis method that comprehensively considers the defect type, location, and heat flow blockage degree enables the defect assessment to be directly associated with the core function of the fuse, thereby more accurately judging the actual impact of the defect on the product performance.

[0095] Please continue to refer to Figure 1 , according to the interface defect suspected point distribution map, apply a micro-pulse thermal excitation to the fuse element, capture the thermal pulse propagation process and the temperature sensor response, and generate an interface thermal resistance distribution map;

[0096] In an embodiment of the present invention, the applying a micro-pulse thermal excitation to the fuse element according to the interface defect suspected point distribution map, capturing the thermal pulse propagation process and the temperature sensor response, and generating an interface thermal resistance distribution map includes:

[0097] According to the interface defect suspected point distribution map, determine the critical path through which the heat flow passes, set a thermal excitation point array on the critical path, increase the density of thermal excitation points in the area near the microbubble position in the thermal excitation point array, adjust the power parameter of the thermal excitation points near the uneven filler distribution area, and extend the pulse duration of the thermal excitation points passing through the microcrack area;

[0098] Apply a micro-pulse thermal excitation point by point to the thermal excitation point array corresponding to the fuse element;

[0099] For each power thermal pulse of each thermal excitation point, collect the spatio-temporal evolution data of the temperature field, record the temperature distribution, propagation speed, and attenuation characteristics of the thermal pulse from the fuse element to the temperature sensor, and generate a temperature field evolution data set;

[0100] Synchronously collect the electrical signal response curve generated by the temperature sensor for each power thermal pulse of each thermal excitation point, calculate the temperature-electrical signal transfer function according to the temperature field evolution data set and the electrical signal response curve, and obtain the sensor response characteristic parameters;

[0101] Perform Fourier thermal wave analysis on the temperature field evolution data set, and calculate the thermal resistance value and thermal diffusivity of the area corresponding to each thermal excitation point according to the thermal wave amplitude attenuation and phase delay;

[0102] Map the thermal resistance value, thermal diffusivity, and the sensor response characteristic parameters to the spatial coordinate system of the interface area, and perform spatial correlation analysis on the thermal resistance value, thermal diffusivity, and the suspected interface defect point distribution map to generate an interface thermal resistance distribution map including thermal conduction performance and sensor response performance.

[0103] Specifically, determining the key heat flow path and setting the thermal excitation point array according to the suspected interface defect point distribution map is realized by a heat path analysis system. The specific method is as follows: import the suspected interface defect point distribution map; run the heat flow simulation calculation to determine the main heat conduction path from the fuse element to the temperature sensor; arrange the thermal excitation point array on this path, with a basic density of setting one thermal excitation point for every 100 μm × 100 μm area. Differentiated treatment is carried out for the thermal excitation points in different types of defect areas: within a range of 50 μm around the position of the microbubble, the density of the thermal excitation points is doubled (such as increasing from 1 per 100 μm × 100 μm to 4 per 100 μm × 100 μm) to achieve precise measurement of the microbubble boundary; for the thermal excitation points near the area with uneven filler distribution, adjust the power parameter to 0.5 to 1.5 times that of the normal area, because the uneven filler distribution will cause local heat capacity changes and different powers are required to cause the same temperature rise; for the thermal excitation points passing through the microcrack area, extend the pulse duration from the standard 100 μs to 300 μs to fully observe the hindering effect of the crack on heat diffusion. In a 5 mm × 5 mm fuse module, a typical thermal excitation point array contains 50 - 100 thermal excitation points. This customized thermal excitation point array design for different defect types can accurately characterize the thermal effects of various defects, far superior to the traditional uniform excitation method.

[0104] Applying micro-pulse thermal excitation point by point to the thermal excitation point array corresponding to the fusing element is achieved through a precision micro-joule laser system. This system focuses the laser into a tiny spot with a diameter of 30 microns, precisely aligns it with each thermal excitation point, and generates transient thermal pulses. The specific operation is as follows: For each thermal excitation point, thermal pulses of three powers are applied respectively, namely, low-power thermal pulses (5 - 10 mW) corresponding to 0.5 times the rated current, medium-power thermal pulses (10 - 20 mW) corresponding to 1.0 times the rated current, and high-power thermal pulses (20 - 30 mW) corresponding to 1.5 times the rated current. These three power levels respectively simulate the heat generation conditions of the fuse under low load, normal load, and critical overload states. For a thermal excitation point array (including 60 points), it takes about 30 minutes to complete all thermal pulse tests. This multi-power-level thermal pulse design can comprehensively evaluate the influence of interface defects under different working conditions, especially the situation where some defects are not obvious at low power but will significantly affect the heat conduction performance at high power.

[0105] Collecting the spatio-temporal evolution data of the temperature field is completed through a high-speed infrared thermal imaging system. This system includes a micro-region high-resolution infrared detector and high-speed image acquisition equipment, with a spatial resolution of 5 microns / pixel, a temperature resolution of 0.01 °C, and a frame rate of 1000 frames per second. The system records the whole process of the thermal pulse propagating from the generation point to the temperature sensor, including: temperature distribution (reflecting the spatial distribution state of the temperature at each point), propagation speed (the rate at which the heat wave front moves, with the unit of mm / s), and attenuation characteristics (the attenuation law of the temperature peak with distance). For example, in a defect-free sample, the thermal pulse propagates at a speed of about 3 mm / s, and the temperature peak attenuates by about 40% for every 1 mm of propagation; while when the thermal pulse passes through a micro-bubble with a diameter of 30 microns, the local propagation speed drops to 1.5 mm / s, and the temperature peak attenuation increases to 65%. For the three power pulses of each thermal excitation point, 200 - 500 frames of temperature field images are recorded to form a complete temperature field evolution data set. This high spatio-temporal resolution temperature field acquisition method can capture the thermal effects of tiny defects that are difficult to observe by traditional methods, such as asymmetric thermal diffusion caused by micro-cracks and increased local thermal resistance caused by micro-bubbles.

[0106] Synchronous acquisition of the temperature sensor's electrical signal response is achieved using a high-precision data acquisition system. This system records the temperature sensor's output signal waveform at a sampling rate of 10kHz, synchronized with the infrared thermal imaging system. The temperature-to-electrical signal transfer function (TEF) is the mathematical relationship between temperature changes and the sensor's output signal. It is calculated by extracting the temperature-time curve at the sensor location from the temperature field evolution dataset; time-aligning this temperature curve with the sensor's electrical signal-time curve; and fitting the transfer function model of the two using a system identification algorithm (such as the least squares method). Sensor response characteristic parameters include sensitivity (the change in the electrical signal per unit temperature change, expressed in mV / °C), response time (the time required to reach 90% of the final value, expressed in milliseconds), and linearity (the deviation of the response curve from ideal linearity, expressed in percent). For example, in a high-quality sample, the temperature sensor has a sensitivity of 2.5mV / °C and a response time of 5ms, with a linearity deviation of less than 2%. However, when microbubbles are present between the sensor and the fuse element, the response time increases to 12ms, and the linearity deviation increases to 7%. These parameters directly reflect the impact of interface defects on the temperature sensing function and provide an important basis for evaluating the reliability of the fuse protection function.

[0107] Fourier thermal wave analysis of the temperature field evolution dataset is performed using thermal wave analysis software. Fourier thermal wave analysis is a data processing method based on heat conduction theory. The specific steps are: Fourier transform the temperature-time curve of each spatial point to obtain the amplitude and phase of different frequency components; according to heat conduction theory, the amplitude decay rate of the thermal wave is related to the thermal diffusivity of the material, and the phase delay is related to the thermal resistance; by solving the heat conduction equation, the thermal resistance of each area (unit: K·cm) is calculated. 2 / W) and thermal diffusivity (in mm 2 / s). In actual analysis, the thermal resistance of a defect-free area is about 0.5K·cm 2 / W, thermal diffusion coefficient is about 0.2mm 2 / s; and the thermal resistance of the area containing microbubbles can be as high as 2.0K·cm 2 / W, thermal diffusion coefficient is reduced to 0.05mm 2 This Fourier thermal wave-based analysis method can extract rich frequency-domain information from time-domain data and is particularly suitable for analyzing complex systems with heat conduction processes at multiple time scales, such as the interface region of a micro-intelligent fuse.

[0108] Mapping the thermal resistance value, thermal diffusivity, and sensor response characteristic parameters to a spatial coordinate system and performing spatial correlation analysis is achieved through a three-dimensional visualization system. The specific steps are as follows: Establish a three-dimensional coordinate system centered on the interface between the fuse element and the temperature sensor; map the thermal resistance value and thermal diffusivity data to this coordinate system according to their spatial positions to form a thermal characteristic distribution map; map the sensor response characteristic parameters to the same coordinate system according to the corresponding regions; perform spatial alignment and correlation analysis on the above distribution map and the suspected interface defect point distribution map to identify the corresponding relationship between thermal characteristic anomalies and specific defects. The finally generated interface thermal resistance distribution map is a multi-layered pseudo-color image, including: The base layer shows the thermal resistance value distribution (high thermal resistance regions are represented by red); the middle layer shows the thermal diffusivity distribution (low diffusivity regions are represented by blue); the top layer shows the sensor response characteristic parameters (such as response delay is marked with yellow); the defect positions are marked with specific symbols (such as microbubbles are represented by circles, and microcracks are represented by line segments). This multi-dimensional information integrated thermal resistance distribution map not only shows the thermal conduction performance distribution in the interface region but also intuitively displays the spatial correlation between defects and thermal performance anomalies, providing an intuitive basis for the hazard assessment of interface defects.

[0109] Please continue to refer to Figure 1 , perform acoustic microscopy analysis based on the thermal resistance anomaly region in the interface thermal resistance distribution map, collect multi-dimensional acoustic parameters, and generate a fine classification map of interface defects;

[0110] In an embodiment of the present invention, the performing acoustic microscopy analysis based on the thermal resistance anomaly region in the interface thermal resistance distribution map, collecting multi-dimensional acoustic parameters, and generating a fine classification map of interface defects includes:

[0111] According to the interface thermal resistance distribution map, identify the region where the thermal resistance value exceeds 30% of the reference value as the key region for acoustic detection, perform acoustic scanning on the key region for acoustic detection, and obtain four basic acoustic parameters: acoustic wave reflection intensity, acoustic wave transmission intensity, sound velocity change, and acoustic wave phase delay;

[0112] Perform correlation analysis between the basic acoustic parameters and the material characteristics of the interface region, analyze the acoustic wave reflection intensity and sound velocity change for the metal-epoxy resin interface region, analyze the acoustic wave transmission intensity and phase delay for the epoxy resin filler region, and analyze the acoustic wave reflection pattern and phase jump characteristics for the temperature sensor-epoxy resin interface region to generate defect acoustic characteristic data;

[0113] Perform pattern recognition on the defect acoustic characteristic data, and determine the defect type, position, size, and severity according to the strong reflection-weak transmission combination characteristics of microbubbles, the medium reflection-nonlinear phase delay combination characteristics of uneven filler distribution, and the directional reflection-sharp phase jump combination characteristics of microcracks, and generate a fine classification map of interface defects.

[0114] Specifically, a thermal resistance analysis system is used to identify key inspection areas based on the interface thermal resistance distribution map. The method involves first determining the baseline thermal resistance value of a defect-free standard sample, which reflects the normal thermal conductivity characteristics of the interface area. Then, the deviation ratio of the thermal resistance value of each region of the sample under test from the baseline value is calculated. Finally, areas with thermal resistance values exceeding 30% of the baseline value are identified as key areas for acoustic inspection. The 30% threshold is selected based on historical experimental data: deviations below 20% are likely normal manufacturing fluctuations, while values above 40% indicate obvious defects. 30% falls into a gray area, requiring further acoustic verification. Acoustic scanning of key inspection areas is performed using an acoustic microscopy system, which emits ultrasonic waves at a frequency of 1 GHz and achieves a spatial resolution of 1.5 microns. The system collects four basic acoustic parameters: acoustic reflection intensity (the intensity of the echo reflected from the interface, measured in decibels), acoustic transmission intensity (the intensity of the sound wave passing through the interface), acoustic velocity variation (the change in the speed of sound waves propagating through different materials), and acoustic phase delay (the phase shift after the sound wave passes through the material). This high-frequency acoustic microscopy technique can detect tiny defects that are difficult to find with traditional methods, and is particularly sensitive to microstructural changes in epoxy resins.

[0115] The material acoustic analysis system correlates basic acoustic parameters with the material properties of the interface region. This system selects the most sensitive combinations of acoustic parameters for analysis based on the material composition characteristics of different interface regions. The metal-epoxy interface region focuses on acoustic reflection intensity and velocity variations, as changes in the interface state significantly affect these two parameters due to the significant difference in acoustic impedance between metal and epoxy. The epoxy filler region focuses on acoustic transmission intensity and phase delay, which are particularly sensitive to changes in filler distribution. The temperature sensor-epoxy interface region focuses on acoustic reflection patterns and phase jump characteristics, as defects at the semiconductor-epoxy interface produce distinctive reflection patterns and phase shifts. Using data processing algorithms, the system extracts key acoustic parameters from each region, generating defect acoustic signature data. For example, in areas of concentrated filler, transmission intensity exhibits irregular attenuation (±3dB fluctuations) and phase delay exhibits nonlinear variations (the phase-distance curve deviates from a straight line). In contrast, in normal regions, transmission intensity decays uniformly, and phase delay exhibits linear variations. This regionalized analysis method fully considers the material diversity in the interface area of the micro smart fuse module and can extract the most valuable characteristic information from complex acoustic data.

[0116] Pattern recognition of the acoustic feature data of defects is accomplished through an acoustic feature classification system. This system automatically identifies and classifies based on the acoustic response characteristics of different types of defects: Microbubbles exhibit a strong reflection - weak transmission combined feature, with the reflection intensity usually 6 - 10 dB higher than the surrounding area and the transmission intensity 10 - 15 dB lower; uneven filler distribution exhibits a medium reflection - non - linear phase delay combined feature, with the reflection intensity varying within the range of 3 - 6 dB and the phase delay curve showing an obvious non - linear characteristic; microcracks exhibit a directional reflection - sharp phase jump combined feature, with the reflection pattern linearly distributed along the crack direction and the phase undergoing a sudden change (>90° jump) at the crack position. The system determines the type, location, size, and severity of each defect through template matching and feature clustering algorithms. The finally generated fine - classification map of interface defects uses multi - level coding: Different colors represent defect types (e.g., red represents microbubbles, blue represents uneven filler distribution, and green represents microcracks); brightness represents severity; the defect contour shows the location and size. For example, in a 5×5 mm micro - intelligent fuse module, the fine - classification map shows 3 microbubbles (with diameters of 15 μm, 25 μm, and 40 μm respectively), 2 areas of uneven filler distribution (with areas of 0.2 mm 2 and 0.15 mm 2 ), and 1 microcrack (with a length of 120 μm). This method of fine - classification of defects based on acoustic features can provide more detailed defect information than the thermal resistance distribution map, especially having significant advantages in terms of defect types and micro - morphology.

[0117] Please continue to refer to Figure 1 , according to the fine - classification map of the interface defects, conduct an electro - thermal transient response co - test on the fuse module, record the temperature rise characteristics of the fusing element, the interface temperature change, and the sensor response under various current conditions, and obtain the test results of performance parameters;

[0118] In an embodiment of the present invention, the conducting an electro - thermal transient response co - test on the fuse module according to the fine - classification map of the interface defects, recording the temperature rise characteristics of the fusing element, the interface temperature change, and the sensor response under various current conditions, and obtaining the test results of performance parameters includes:

[0119] According to the defect types and distribution positions in the fine - classification map of the interface defects, divide the fuse module into a defect - free area, a microbubble - affected area, an uneven - filler - distribution - affected area, and a microcrack - affected area;

[0120] Apply a standard current condition to the defect - free area to obtain reference performance data, apply a current slow - climb condition to the microbubble - affected area, apply a current step condition to the uneven - filler - distribution - affected area, apply a current pulse condition to the microcrack - affected area, and record the temperature rise characteristic curves of the fusing element in each area;

[0121] Collect the interface temperature change data of the defect-free area, the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks under their respective current conditions, calculate the temperature gradients and heat flux distributions of each area, and obtain the dynamic distribution map of the interface temperature;

[0122] Synchronously record the response signals of the temperature sensors in the defect-free area, the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks under their respective current conditions, calculate the temperature transfer efficiency parameters in combination with the temperature rise characteristic curve of the fuse element, and calculate the response time parameters and temperature-signal distortion parameters in combination with the dynamic distribution map of the interface temperature;

[0123] Compare and analyze the reference performance data with the performance data of the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks, establish the correlation between the defect types and the temperature transfer efficiency parameters, response time parameters, and temperature-signal distortion parameters, and obtain the test results of the performance parameters.

[0124] Specifically, the partitioning of the fuse module according to the fine classification map of interface defects is realized by a defect influence area division system. This system first determines the influence ranges of various defects: the influence range of microbubbles is 50 microns outward from the bubble edge; the influence range of uneven filler distribution is 100 microns outward from the uneven boundary; the influence range of microcracks is 75 microns on each side of the crack. Then the module is divided into four areas: the defect-free area (without any defects and more than 150 microns away from the edges of all defects), the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks. In a 5×5 mm micro intelligent fuse sample, the defect-free area accounts for about 60% of the total area, and the areas affected by the three types of defects account for 15%, 18%, and 7% respectively. This area division based on defect types enables subsequent tests to be customized according to the characteristics of different defects, so as to more accurately evaluate the actual impact of various defects on the function of the fuse.

[0125] Applying differential current conditions to different regions is achieved through a precision current control system. This system can generate four typical current conditions: standard current condition (constant at 80% of the rated current for 30 seconds), current slow-rise condition (current rises from 50% of the rated value to 100% at a rate of 2% per second, with a total duration of 25 seconds), current step condition (current suddenly changes from 50% to 90% of the rated value within 0.1 second and lasts for 30 seconds), and current pulse condition (120% rated current pulse lasts for 100 milliseconds and then returns to 60% of the rated value, repeating 5 times). After applying the corresponding current conditions to each region, the temperature rise characteristic curve of the fuse element is recorded through a micro-thermocouple array, that is, the curve of the fuse element temperature changing with time. The selection of these four current conditions is based on an in-depth analysis of the working characteristics of the micro-intelligent fuse: the current slow-rise can most expose the influence of micro-bubbles on continuous heat conduction; the current step can most reflect the influence of uneven filler distribution on heat capacity; the current pulse can most test the influence of micro-cracks on thermal shock response.

[0126] Collecting the data of the interface temperature change is completed through an infrared micro-thermal imaging system. The spatial resolution of this system is 5 microns / pixel, the temperature resolution is 0.01 °C, and the sampling frequency is 100 Hz. At the same time, the temperature changes in four regions under their respective current conditions are recorded. According to the collected temperature data, the temperature gradient (unit: °C / mm) and heat flux distribution (unit: W / cm 2 ) of each region are calculated. The temperature gradient refers to the rate of change of temperature in space, which is obtained by dividing the temperature difference between adjacent measurement points by the distance; the heat flux distribution is calculated according to Fourier's law of heat conduction (heat flux = -k · temperature gradient, where k is the thermal conductivity). The system synthesizes the calculation results into a dynamic distribution map of the interface temperature and displays the whole process of heat transfer from the fuse element to the temperature sensor in the form of a pseudo-color dynamic video. In the test of the region with uneven filler distribution, the temperature gradient shows obvious non-uniformity, and the ratio of the maximum value to the minimum value can reach 3:1, while in the defect-free region, this ratio is usually less than 1.5:1. This microscopic-scale heat distribution analysis can intuitively reveal the disturbance effect of defects on the heat conduction path.

[0127] Synchronously recording the response signal of the temperature sensor is achieved through a high-precision data acquisition system. The sampling rate of this system is 10 kHz, and the precision is 16 bits. It synchronously records the electrical signal output of the sensor under various conditions in each area. Based on the collected data, the temperature transfer efficiency parameter (the ratio of the temperature measured by the sensor to the actual temperature of the fusing element, with an ideal value of 1.0) is calculated by combining the temperature rise characteristic curve of the fusing element; the response time parameter (the time required for the sensor signal to reach 90% of the final value, in milliseconds) and the temperature-signal distortion parameter (the deviation of the sensor output signal from the ideal linear response, in percentage) are calculated by combining the dynamic distribution map of the interface temperature. During the test, the temperature transfer efficiency in the defect-free area is usually between 0.95 and 0.98, the response time is 3 to 5 milliseconds, and the distortion is less than 2%; while in the area affected by microbubbles, the temperature transfer efficiency drops to 0.75 to 0.85, the response time extends to 8 to 12 milliseconds, and the distortion increases to 5 to 8%. These three parameters comprehensively reflect the functional performance of the temperature sensor and are directly related to the protection function reliability of the micro intelligent fuse.

[0128] Comparing and analyzing the reference performance data with the performance data of each defect area is achieved through a performance correlation analysis system. This system first calculates the deviation rate of the performance parameters of each defect area from the reference value; then establishes the mapping relationship between the defect type and the performance parameters; finally generates a test result report of the performance parameters. The test results show that microbubbles mainly affect the temperature transfer efficiency, with an average reduction of 20%; uneven distribution of fillers mainly affects the signal distortion, increasing by 3 to 7%; microcracks mainly affect the response time, extending by 40 to 60%. This defect-performance correlation analysis reveals the differential impact mechanism of different types of defects on the function of the micro intelligent fuse, providing a function-oriented scientific basis for subsequent defect hazard assessment and quality grading. Compared with the traditional detection method that only focuses on the defect itself, this function-oriented test method can better reflect the impact of defects on the actual performance of the product and has higher practical value.

[0129] Please continue to refer to Figure 1 , according to the test results of the performance parameters, apply targeted accelerated aging treatment to the samples with different types of interface defects, conduct periodic detections during the aging process, track the defect evolution process and the performance degradation trend, and determine the quality grade and reliability prediction result.

[0130] In an embodiment of the present invention, the applying targeted accelerated aging treatment to the samples with different types of interface defects according to the test results of the performance parameters, conducting periodic detections during the aging process, tracking the defect evolution process and the performance degradation trend, and determining the quality grade and reliability prediction result includes:

[0131] According to the degree of influence of defect types on performance in the test results of the performance parameters, temperature cycle aging is applied to the samples affected by microbubbles, high-temperature and high-humidity aging is applied to the samples affected by uneven filler distribution, and temperature-current composite cycle aging is applied to the samples affected by microcracks to generate an aging sample set;

[0132] At key nodes during the aging process of the aging sample set, interface thermal resistance imaging, acoustic microscopy analysis, and electro-thermal transient response tests are carried out to obtain defect size change data, interface structure change data, and performance parameter change data;

[0133] Correlation analysis is carried out on the defect size change data, interface structure change data, and performance parameter change data to calculate the defect size growth rate and the hazard coefficients of various types of defects;

[0134] According to the defect size growth rate and the hazard coefficients of various types of defects, the micro intelligent fuse module is divided into four quality grades: A, B, C, and D. The expected life of the samples in each grade is calculated to determine the quality grade and the reliability prediction result.

[0135] Specifically, the targeted aging treatment of samples with different types of defects is achieved through a multi-mode accelerated aging system. This system designs differentiated aging schemes according to the physical characteristics of different defect types: temperature cycle aging (-40°C to +125°C, 500 cycles, heating and cooling rate 3°C / minute) is applied to the samples affected by microbubbles because temperature changes can cause gas expansion and contraction, accelerating the change in bubble size; high-temperature and high-humidity aging (85°C / 85% relative humidity, lasting 1000 hours) is applied to the samples affected by uneven filler distribution, and moisture can promote the hydrolysis reaction at the interface between the filler and the resin, changing the filler distribution state; temperature-current composite cycle aging (temperature -20°C to +100°C, current changing synchronously from 50% to 90% of the rated value, 300 cycles) is applied to the samples affected by microcracks. The combination of thermal stress and mechanical stress generated by the current can effectively expand microcracks. In a batch of typical test samples, 20 samples are processed with each of the three aging schemes to form a complete aging sample set. This customized aging scheme design greatly improves the efficiency and pertinence of the aging test.

[0136] Multidimensional detection of the aging sample set at key nodes during the aging process is accomplished through an automated node detection system. The system performs three tests on the samples at five key nodes of 0%, 20%, 50%, 80%, and 100% during the aging process: interfacial thermal resistance imaging (detecting changes in heat conduction), acoustic microscopy analysis (detecting changes in microstructure), and electro-thermal transient response testing (detecting changes in functional performance). The data obtained through these detections include: defect size change data (such as a 10% increase in the diameter of microbubbles and a 25% increase in the length of microcracks), interfacial structure change data (such as a 15% expansion in the area of filler aggregation regions), and performance parameter change data (such as an 8% decrease in temperature transfer efficiency and a 20% increase in response time). In a test where a microbubble affects the sample, after 500 temperature cycles, the bubble diameter increases from an initial 30 microns to 42 microns, the thermal resistance value increases by 45%, and the temperature transfer efficiency decreases by 15%. This multi-dimensional dynamic detection method can comprehensively capture the entire process of defect evolution, far superior to the traditional method that only detects the states before and after aging.

[0137] Correlation analysis of multi-dimensional change data is achieved through a defect evolution analysis system. The system first calculates the defect size growth rate, that is, the change rate of the defect size with aging time (such as the growth rate of the microbubble diameter is 4% / 100 cycles); then, according to the relationship between performance parameter degradation and defect size growth, it calculates the hazard coefficient of various defects, which reflects the degree of performance degradation caused by a unit increase in defect size. The test data shows that the hazard coefficient of microbubbles is 0.35 (that is, when the bubble diameter increases by 10%, the temperature transfer efficiency drops by 3.5%); the hazard coefficient of uneven filler distribution is 0.25; the hazard coefficient of microcracks is the highest, reaching 0.6. This correlation analysis method of defect evolution and performance degradation reveals the hazard mechanisms and severity of different defects, providing a quantitative basis for scientifically evaluating product reliability.

[0138] Quality grading based on the defect size growth rate and hazard coefficient is completed through a reliability assessment system. The system classifies products into four grades according to the comprehensive score (defect size growth rate × hazard coefficient × 100): Grade A (comprehensive score < 10, all performance parameters are stable, expected life > design life × 1.5), Grade B (comprehensive score 10 - 20, performance parameter fluctuation < 10%, expected life ≈ design life), Grade C (comprehensive score 20 - 30, performance parameter fluctuation 10% - 20%, expected life is 0.7 - 0.9 times the design life), and Grade D (comprehensive score > 30, performance parameter fluctuation > 20%, expected life < 0.7 times the design life). This quality grading method based on the dynamic characteristics of defect evolution makes product grading more scientific and predictable, and can accurately predict the long-term reliability of products in actual applications.

[0139] Another embodiment of the present invention provides a highly integrated micro intelligent fuse module, and the preparation process of the highly integrated micro intelligent fuse module adopts the preparation process of the highly integrated micro intelligent fuse module described in any of the above embodiments.

[0140] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structural transformation made by using the content of the specification and drawings of the present invention under the inventive concept of the present invention, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A preparation process of a highly integrated micro intelligent fuse module, characterized in that, Including: Performing high-frequency electromagnetic disturbance detection on the interface region between the fusing element and the temperature sensor of the fuse module to obtain the electromagnetic response fingerprint of the interface region, and determining the suspected interface defect point distribution map based on the electromagnetic response fingerprint; According to the suspected interface defect point distribution map, applying a micro-pulse thermal excitation to the fusing element, capturing the thermal pulse propagation process and the temperature sensor response, and generating an interface thermal resistance distribution map; According to the thermal resistance abnormal region in the interface thermal resistance distribution map, performing acoustic microscopy analysis, collecting multi-dimensional acoustic parameters, and generating an interface defect fine classification map; According to the interface defect fine classification map, performing an electro-thermal transient response collaborative test on the fuse module, recording the temperature rise characteristics of the fusing element, the interface temperature change and the sensor response under various current conditions, and obtaining the performance parameter test results; According to the performance parameter test results, applying targeted accelerated aging treatment to samples with different types of interface defects, periodically detecting during the aging process, tracking the defect evolution process and the performance degradation trend, and determining the quality grade and reliability prediction results.

2. The preparation process of the highly integrated micro intelligent fuse module according to claim 1, characterized in that, The performing high-frequency electromagnetic disturbance detection on the interface region between the fusing element and the temperature sensor of the fuse module to obtain the electromagnetic response fingerprint of the interface region, and determining the suspected interface defect point distribution map based on the electromagnetic response fingerprint includes: Applying an electromagnetic disturbance with a first frequency in the range of 2.0 - 3.0 GHz to the fusing element region of the fuse module, and applying an electromagnetic disturbance with a second frequency in the range of 5.0 - 6.0 GHz to the temperature sensor region; Capturing the electromagnetic reflection wave and transmission wave of the interface region between the fusing element and the temperature sensor, recording the reflection coefficient and phase change of the electromagnetic reflection wave, and the transmission coefficient and phase delay of the transmission wave; Performing multi-parameter fusion analysis on the reflection coefficient, phase change, transmission coefficient and phase delay to generate the electromagnetic response fingerprint of the interface region; Comparing the electromagnetic response fingerprint with the standard reference value, and calculating the anomaly index according to the deviation degree; According to the anomaly index and the characteristic parameters in the electromagnetic response fingerprint, identifying the characteristic patterns corresponding to microbubbles, uneven filler distribution and microcracks, and determining the suspected interface defect point distribution map.

3. The preparation process of the highly integrated micro intelligent fuse module according to claim 2, characterized in that, The performing multi-parameter fusion analysis on the reflection coefficient, phase change, transmission coefficient and phase delay to generate the electromagnetic response fingerprint of the interface region includes: Dividing the interface region between the fusing element and the temperature sensor into three functional structure regions: metal-epoxy resin transition region, epoxy resin main body region, and semiconductor-epoxy resin transition region, and respectively extracting the reflection coefficient, phase change, transmission coefficient and phase delay for each functional structure region; Based on the distribution characteristics of nano-thermal conductive fillers in epoxy resin, performing a composite analysis on the reflection coefficient and phase change of the metal-epoxy resin transition region to generate metal interface characteristic parameters; Based on the dielectric characteristics of epoxy resin, performing a correlation analysis on the transmission coefficient and phase delay of the epoxy resin main body region to generate filler distribution characteristic parameters; Based on the interfacial characteristics of semiconductor materials and epoxy resins, the reflection coefficient and transmission coefficient of the semiconductor-epoxy resin transition region are compared and analyzed to generate sensor interface characteristic parameters; Integrate and map the metal interface characteristic parameters, filler distribution characteristic parameters, and sensor interface characteristic parameters to generate an electromagnetic response fingerprint of the interface region.

4. The preparation process of the highly integrated micro intelligent fuse module according to claim 2, characterized in that, According to the anomaly index and the characteristic parameters in the electromagnetic response fingerprint, identify the characteristic patterns corresponding to microbubbles, uneven filler distribution, and microcracks, and determine the suspected interface defect point distribution map, including: Conduct a fusing heat flow transfer path analysis on the anomaly index, and divide the interface region into a key heat transfer region, a secondary heat transfer region, and a non-heat transfer region. Among them, the region where the fusing element is in direct contact with the temperature sensor and the distance is less than 100 microns is defined as the key heat transfer region, the region with a distance of 100 - 300 microns from the edge of the key heat transfer region is defined as the secondary heat transfer region, and the region with a distance greater than 300 microns from the edge of the key heat transfer region is defined as the non-heat transfer region; Calculate the covariance matrix of the phase delay and reflection coefficient of each region in the electromagnetic response fingerprint, extract the eigenvalue distribution pattern from the covariance matrix, and determine the position of microbubbles by identifying the bimodal feature in the eigenvalue distribution pattern; Calculate the spatial gradient of the transmission coefficient in the electromagnetic response fingerprint, analyze the anisotropy difference of the spatial gradient, and determine the uneven filler distribution region according to the region where the value of the anisotropy difference is greater than the preset threshold; Perform a differential operation on the electromagnetic responses at the first frequency and the second frequency to obtain a difference signal, extract the linear discontinuity feature in the difference signal, and determine the position and orientation of microcracks according to the linear discontinuity feature; According to the distribution of the microbubble positions, the uneven filler distribution regions, the microcrack positions and orientations in the key heat transfer region, the secondary heat transfer region, and the non-heat transfer region, calculate the hazard level value in combination with the degree of blocking of the heat flow path, and map the microbubble positions, the uneven filler distribution regions, the microcrack positions and orientations, and the hazard level value to the interface region coordinate system to generate a suspected interface defect point distribution map.

5. The preparation process of the highly integrated micro intelligent fuse module according to claim 1, characterized in that, According to the suspected interface defect point distribution map, apply a micro-pulse thermal excitation to the fusing element, capture the heat pulse propagation process and the temperature sensor response, and generate an interface thermal resistance distribution map, including: According to the suspected interface defect point distribution map, determine the key path through which the heat flow passes, set a thermal excitation point array on the key path, increase the density of thermal excitation points in the region near the microbubble position in the thermal excitation point array, adjust the power parameter of the thermal excitation points near the uneven filler distribution region, and extend the pulse duration of the thermal excitation points passing through the microcrack region; Apply a micro-pulse thermal excitation point by point to the thermal excitation point array corresponding to the fusing element; Collect the spatio-temporal evolution data of the temperature field for each power thermal pulse of each thermal excitation point, record the temperature distribution, propagation speed, and attenuation characteristics of the heat pulse propagating from the fusing element to the temperature sensor, and generate a temperature field evolution data set; Synchronously collect the electrical signal response curves generated by the temperature sensors for each power thermal pulse at each thermal excitation point, calculate the temperature-electrical signal transfer function according to the temperature field evolution data set and the electrical signal response curves, and obtain the sensor response characteristic parameters; Perform Fourier thermal wave analysis on the temperature field evolution data set, and calculate the thermal resistance value and thermal diffusivity of the corresponding area of each thermal excitation point according to the thermal wave amplitude attenuation and phase delay; Map the thermal resistance value, thermal diffusivity and the sensor response characteristic parameters to the spatial coordinate system of the interface area, perform spatial correlation analysis on the thermal resistance value, thermal diffusivity and the suspected interface defect point distribution map, and generate an interface thermal resistance distribution map including thermal conduction performance and sensor response performance.

6. The preparation process of the highly integrated micro intelligent fuse module according to claim 1, characterized in that, Perform acoustic microscopy analysis according to the thermal resistance abnormal area in the interface thermal resistance distribution map, collect multi-dimensional acoustic parameters, and generate a fine classification map of interface defects, including: According to the interface thermal resistance distribution map, identify the area where the thermal resistance value exceeds 30% of the reference value as the key area for acoustic detection, perform acoustic scanning on the key area for acoustic detection, and obtain four basic acoustic parameters: acoustic wave reflection intensity, acoustic wave transmission intensity, sound velocity change and acoustic wave phase delay; Perform correlation analysis on the basic acoustic parameters and the material characteristics of the interface area, analyze the acoustic wave reflection intensity and sound velocity change for the metal-epoxy resin interface area, analyze the acoustic wave transmission intensity and phase delay for the epoxy resin filler area, and analyze the acoustic wave reflection mode and phase jump characteristics for the temperature sensor-epoxy resin interface area to generate defect acoustic characteristic data; Perform pattern recognition on the defect acoustic characteristic data, and determine the defect type, position, size and severity according to the strong reflection-weak transmission combination characteristics of microbubbles, the medium reflection-nonlinear phase delay combination characteristics of uneven filler distribution, and the directional reflection-sharp phase jump combination characteristics of microcracks, and generate a fine classification map of interface defects.

7. The preparation process of the highly integrated micro intelligent fuse module according to claim 1, characterized in that, Perform electro-thermal transient response co-test on the fuse module according to the fine classification map of interface defects, record the temperature rise characteristics of the fuse element, interface temperature change and sensor response under various current conditions, and obtain the test results of performance parameters, including: According to the defect type and distribution position in the fine classification map of interface defects, divide the fuse module into defect-free area, microbubble influence area, uneven filler distribution influence area and microcrack influence area; Apply the standard current condition to the defect-free area to obtain the reference performance data, apply the current slow climb condition to the microbubble influence area, apply the current step condition to the uneven filler distribution influence area, apply the current pulse condition to the microcrack influence area, and record the temperature rise characteristic curves of the fuse elements in each area; Collect the interface temperature change data of the defect-free area, microbubble influence area, uneven filler distribution influence area and microcrack influence area under their respective current conditions, calculate the temperature gradient and heat flow distribution of each area, and obtain the dynamic distribution map of interface temperature. Synchronously record the temperature sensor response signals of the defect-free area, the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks under their respective current conditions, calculate the temperature transfer efficiency parameter in combination with the temperature rise characteristic curve of the fusing element, and calculate the response time parameter and the temperature-signal distortion parameter in combination with the dynamic interface temperature distribution map; Compare and analyze the reference performance data with the performance data of the area affected by microbubbles, the area affected by uneven filler distribution, and the area affected by microcracks, establish the correlation between the defect type and the temperature transfer efficiency parameter, the response time parameter, and the temperature-signal distortion parameter, and obtain the test results of the performance parameters.

8. The preparation process of the highly integrated micro intelligent fuse module according to claim 1, characterized in that According to the test results of the performance parameters, apply targeted accelerated aging treatment to the samples with different types of interface defects, perform periodic detection during the aging process, track the defect evolution process and the performance degradation trend, and determine the quality grade and the reliability prediction result, including: According to the degree of influence of the defect type on the performance in the test results of the performance parameters, apply temperature cycle aging to the samples affected by microbubbles, apply high temperature and high humidity aging to the samples affected by uneven filler distribution, and apply temperature-current composite cycle aging to the samples affected by microcracks to generate an aging sample set; Perform interface thermal resistance imaging, acoustic microscopy analysis, and electro-thermal transient response testing on the key nodes of the aging sample set during the aging process to obtain defect size change data, interface structure change data, and performance parameter change data; Perform correlation analysis on the defect size change data, the interface structure change data, and the performance parameter change data, and calculate the defect size growth rate and the hazard coefficients of various defects; According to the defect size growth rate and the hazard coefficients of various defects, classify the micro intelligent fuse module into four quality grades: A, B, C, and D, calculate the expected life of each grade of samples, and determine the quality grade and the reliability prediction result.

9. A highly integrated micro intelligent fuse module, characterized in that, The preparation process of the highly integrated micro intelligent fuse module adopts the preparation process of the highly integrated micro intelligent fuse module according to any one of claims 1 to 8.

Images