Detection Method and System for ESD Devices Used in FAN-OUT Process

Abstract

The present application relates to the technical field of electronic device detection, and particularly to a detection method and system for an ESD device used in a FAN-OUT process. The method includes: performing independent component analysis on the first current data between the ESD protector and the protected device for all pulse corresponding time periods to obtain independent components and their independent weights in each time period; analyzing the falling characteristics of the second current data between the protector and the power supply in each time period, analyzing the similarity characteristics between each independent component and the falling part of the second current data in each time period to obtain the parasitic correlation of each independent component; combining the width of each pulse current to obtain the pulse correlation of each independent component, correcting the parasitic correlation, and combining the independent weights of each independent component in each time period to reconstruct the first current data; and detecting the ESD device. The present application aims to reduce the measurement error and misjudgment risk caused by parasitic effects.

Description

Technical Field

[0001] This application relates to the technical field of electronic device detection, and particularly to a method and system for detecting ESD devices used in the FAN-OUT process. Background Art

[0002] With the continuous development of semiconductor packaging technology, as a new packaging method, FAN-OUT packaging has been widely used in various high-performance electronic products, especially in the fields of mobile phones, computers, automotive electronics, consumer electronics, etc. By extending the pins of the chip to a larger area through the Redistribution Layer (RDL), FAN-OUT packaging not only improves the integration and electrical performance but also enhances the thermal management and heat dissipation performance. However, with the increase in the complexity of these products, electromagnetic compatibility problems, electrostatic discharge (ESD) protection problems, etc. in circuit design have become increasingly prominent. ESD events may cause damage to sensitive components in the circuit. Therefore, it is particularly important to detect the effectiveness of ESD protection devices in FAN-OUT packaging.

[0003] In the electronic packaging of the FAN-OUT process, due to the complexity of circuit design, the influence of parasitic inductance and capacitance has become the main technical challenge in testing ESD protection devices. The signal paths in FAN-OUT packaging are long and dense, resulting in large parasitic inductance and capacitance in the circuit. These parasitic effects may cause serious signal distortion in transient pulse tests. When testing ESD protection devices, these parasitic components will interfere with the current waveform, causing waveform distortion or delay, thus affecting the accurate evaluation of the parameters of ESD protection devices. For example, the current and voltage waveforms may exhibit excessive oscillations during actual discharge due to the presence of parasitic inductance and capacitance, making it difficult to accurately judge the working state of ESD protection devices. Especially in high-frequency and high-speed response transient events, these parasitic effects will lead to misjudgment and affect the reliability evaluation of ESD protection devices. Therefore, it is necessary to reduce the influence of parasitic inductance and capacitance on detection to ensure the accuracy of test results. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method and system for detecting ESD devices used in the FAN-OUT process to solve the above problems.

[0005] In the first aspect of this application, a method for detecting an ESD device used in the FAN-OUT process is provided, and the method includes:

[0006] Obtain the first current data between the ESD protector and the protected device; obtain the second current data between the protected device and the power supply;

[0007] Interpolate the first current data and the second current data for each time period respectively based on the number of data in the longest time period corresponding to the pulse; perform independent component analysis on the first current data for all time periods corresponding to the pulses to obtain a preset number of independent components of the first current data and their independent weights in each time period;

[0008] Analyze the falling characteristics of the second current data for each time period to obtain the falling part of the second current data for each time period, and analyze the similarity characteristics between each independent component of the first current data and the falling part of the second current data for each time period to obtain the parasitic correlation of each independent component;

[0009] Based on the independent component decomposition results of the first current data in the time period corresponding to each pulse, combined with the width of the current of each pulse, obtain the pulse correlation of each independent component, and correct the parasitic correlation; based on the corrected parasitic correlation of each independent component and the independent weight of each independent component in each time period, reconstruct the first current data;

[0010] Detect the ESD device based on the deviation between the reconstructed first current data and the current data under the preset voltage.

[0011] Among them, the specific process of obtaining the falling part of the second current data for each time period is as follows:

[0012] Denote the sequence composed of the interpolated second current data for each time period as the second current data sequence; obtain the second-order difference sequence of the second current data sequence;

[0013] Based on the element distribution characteristics in the second-order difference sequence, obtain the segmentation moment, divide the second current data of the corresponding time period into two parts, and denote the later segmentation part as the falling part.

[0014] Among them, the segmentation moment is specifically the moment corresponding to the minimum minimum value point in the second-order difference sequence.

[0015] Among them, the specific process of obtaining the parasitic correlation of each independent component is as follows:

[0016] Obtain the subsequence of each independent component of the first current data after the segmentation moment corresponding to each time period; obtain the similarity measure between the subsequence obtained for each independent component in each time period and the falling part of the second current data of the corresponding time period; take the mean value of the similarity measures obtained for each independent component in all time periods as the parasitic correlation of each independent component.

[0017] Among them, the specific process of obtaining the pulse correlation of each independent component is as follows:

[0018] Take the independent weight of each independent component in each time period as the numerator; take the time length of each time period as the denominator; extract the dispersion degree of the corresponding scores of each independent component in all time periods, and take the result of the negative correlation mapping of the dispersion degree as the pulse correlation of each independent component.

[0019] Among them, the correction of the parasitic correlation is specifically: take the normalized value of the pulse correlation of each independent component and the positive fusion result of the parasitic correlation as the corrected parasitic correlation of each independent component.

[0020] Among them, the reconstruction of the first current data is specifically:

[0021] Based on the corrected parasitic correlation of each independent component, correct the independent weight of each independent component in each time period to obtain a corrected weight;

[0022] Take the linear combination of each independent component and its corrected weight in each time period as the reconstructed first current data in each time period.

[0023] Among them, the correction process of the independent weight of each independent weight in each time period is: normalize the opposite number of the corrected parasitic correlation of each independent component, and take the product of the obtained normalized result and the independent weight of each independent component in each time period as the corrected weight of each independent component in each time period.

[0024] Among them, the detection of the ESD device is specifically:

[0025] Record the current data under the preset voltage as the static current;

[0026] Calculate the maximum deviation between the reconstructed first current data and the static current; when the maximum deviation is greater than the preset value, it is determined that the ESD device has a problem; otherwise, it is determined that the ESD device has no problem.

[0027] In a second aspect, an embodiment of the present application further provides a detection system for an ESD device used in a FAN-OUT process, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0028] The above solution obtains the current data under the transmission line pulse, decomposes the current data between the ESD protector and the protected device, fully considers the influence of parasitic effects, effectively suppresses the components greatly affected by parasitic effects, thereby improving the accuracy of detection. Specifically, in this application, the current data between the ESD protector and the protected device is decomposed, the parasitic correlation between each independent component and the operating current of the protected circuit is calculated, and combined with the strength characteristics of parasitic effects under different pulse widths, the parasitic correlation is corrected, and then more realistic current data is reconstructed. Based on the optimized current data, this application can more accurately evaluate the key parameters of the ESD protection device and ensure the reliability of the detection results.

[0029] In addition, by testing the protection ability and tolerance of the ESD device under different pulse current conditions, this application provides a systematic and standardized test process, which helps to reduce the measurement error and misjudgment risk caused by parasitic effects, and improve the ESD protection optimization efficiency in the product design stage. Finally, this method not only improves the performance evaluation accuracy of the ESD protection device, but also reduces the later repair and rework costs, enhances the overall reliability and market competitiveness of the product, and has important engineering application value for ESD protection detection in the FAN-OUT process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of the steps of a method for detecting an ESD device for the FAN-OUT process provided by an embodiment of this application;

[0031] Figure 2 It is a schematic flowchart for reconstructing the first current data provided by an embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] In the description of the embodiments of this application, words such as "exemplary", "or", "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary", "or", "for example" is intended to present related concepts in a specific manner.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0034] In addition, it should be noted that the terms "first" and "second" in this application and the accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. In the methods disclosed in the embodiments of this application or the methods shown in the flowcharts, which include one or more steps for implementing the methods, without departing from the scope of protection of this application, the execution order of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0036] The following specifically describes the specific solutions of the detection method and system for ESD devices for FAN-OUT processes provided by this application with reference to the accompanying drawings.

[0037] Please refer to Figure 1 , which shows the flowchart of the steps of the detection method for ESD devices for FAN-OUT processes provided by an embodiment of this application. The method includes the following steps:

[0038] The first step: Obtain the first current data between the ESD protector and the protected device; obtain the second current data between the protected device and the power supply.

[0039] In an actual circuit, due to its complex wiring and high-density integration characteristics, the FAN-OUT package is prone to introducing relatively large parasitic inductance and parasitic capacitance. These non-ideal factors will have a significant impact on the performance of the circuit. The parasitic capacitance and inductance during the operation of the protected device in the FAN-OUT package will affect the propagation path and energy distribution of the ESD event, and thus affect the leakage characteristics of the ESD protection device.

[0040] This application considers performing real-time detection of ESD devices through the change data of the leakage current during the operation of the protected device in the FAN-OUT package. Here, the same rise time of 1 ns, fall time of 100 ns, and maximum voltage of 20 kV are set, and the fast rise time of the standard ESD waveform is adopted, which complies with the IEC 61000-4-2 standard; then, the response of the ESD device is tested by changing the duration of the pulse current; among them, the duration of each pulse starts from 20 ns and increases by 5 ns in sequence until it reaches 100 ns; this method can help evaluate the protection ability and tolerance of the ESD device under different pulse conditions.

[0041] Perform current detection between the ESD protector and the protected device to obtain the magnitude of the leakage current, denoted as the first current data ; then, current detection is performed between the protected device and the power supply to analyze the operating conditions of the protected circuit at different times, which is recorded as the second current data .

[0042] The second step: Based on the number of data in the longest time period corresponding to the pulse, interpolation processing is respectively performed on the first current data and the second current data in each time period; independent component analysis is performed on the first current data in all time periods corresponding to the pulses to obtain a preset number of independent components of the first current data and their independent weights in each time period.

[0043] Due to its complex wiring and high-density integration characteristics, the FAN-OUT package is prone to introducing a large parasitic component inductance. The parasitic effect may cause a higher leakage current to be detected after an ESD event, but this may not be the failure of the ESD device, but a measurement error caused by the influence of the parasitic effect.

[0044] Therefore, in this application, the first current data is decomposed by means of independent component analysis (ICA) Through the association between each component and the parasitic effect, the influence of the parasitic effect on the first current data is removed.

[0045] Before performing ICA decomposition, it is necessary to segment the current change data according to the change of the pulsed current applied in different time periods Specifically: According to the start time and end time of the pulsed current generated by each pulse generator, the time period corresponding to each pulsed current is obtained as the time period corresponding to each pulsed current.

[0046] In order to analyze the performance of the ESD protector under different pulsed currents, since the pulse duration corresponding to each pulsed current is different, the first current data in the r-th time period is denoted as , in order to analyze the situation affected by the parasitic effect in different time periods, it is necessary to perform interpolation processing on the first current data in different time periods, obtain the number of first current data in the longest time period corresponding to the pulsed current, and interpolate the first current data in the remaining time periods in turn, so that the data lengths of the first current data in all time periods are the same. In this embodiment, cubic spline interpolation is used for data filling, and the filled first current data in the r-th time period is denoted as . Among them, cubic spline interpolation is a well-known prior art, and this application will not elaborate on it. The interpolated first current data in the r-th time period is denoted as ;Use the first current data after interpolation for all time periods as the input of the ICA algorithm, preset the number of independent components, and output the preset number of independent components and the independent weights of each independent component in each time period; among them, the ICA algorithm is a well-known existing technology, and this application will not elaborate on it in detail.

[0047] It should be understood that the number of data points of the first current data for all time periods after interpolation is the same; the number of elements of each independent component is the same as the number of the first current data after interpolation for each time period.

[0048] Correspondingly, in the time period corresponding to each pulse, interpolate the second current data between the protected device and the power supply according to the same interpolation method as the first current data, so that the data lengths of the second current data in all time periods are the same, and denote the second current data after interpolation for every r time periods as 。

[0049] The third step: Analyze the decline characteristics of the second current data for each time period to obtain the declining part of the second current data for each time period, and analyze the similarity characteristics between each independent component of the first current data and the declining part of the second current data for each time period to obtain the parasitic correlation of each independent component.

[0050] During an ESD event, due to the existence of parasitic capacitance in the protected circuit, the second current data will rise rapidly and then gradually decline. When the second current data rises rapidly, the parasitic capacitance will be charged, but its detection of the first current data is relatively small; while in the stage when the second current data gradually declines, the transient impact of the ESD event begins to disappear, the second current at the power supply end gradually returns to normal, but the parasitic capacitance will gradually release the stored charge, making the current decline slowly.

[0051] This application analyzes the decline characteristics of the second current data between the power supply and the protected device, analyzes its similarity characteristics with the independent components of the first current data, and calculates the parasitic correlation of each independent component.

[0052] Specifically, denote the sequence composed of the second current data after interpolation for each time period as the second current data sequence; obtain the second-order difference sequence of the second current data sequence; when the second current data in the corresponding time period changes from rising to gradually declining, the second-order difference value usually shows a local minimum value. Therefore, take the moment corresponding to the smallest minimum value point in the second-order difference sequence as the segmentation moment, divide the second current data in the corresponding time period into two parts, denote the later segmentation part as the declining part, and the remaining part as the rising part.

[0053] Further, according to the similarity between the descending part of the second current data in each time period and each independent component of the first current data, the parasitic correlation of each independent component is obtained. The specific steps are as follows: Obtain the subsequences of each independent component of the first current data after the corresponding segmentation moment in each time period; Obtain the similarity measure between the subsequences obtained by each independent component in each time period and the descending part of the second current data in the corresponding time period; Take the mean value of the similarity measures obtained by each independent component in all time periods as the parasitic correlation of each independent component. In this embodiment, the calculation method of the Pearson correlation coefficient is used for the similarity measure between sequences.

[0054] It should be understood that when the correlation between the subsequences obtained by each independent component for each pulse and the descending part of the second current data in the corresponding time period is relatively high, under the action of multiple pulsed currents, the greater the influence of the parasitic effect on this independent component.

[0055] The fourth step: According to the independent component decomposition results of the first current data in the corresponding time period of each pulse, combined with the width of each pulse current, obtain the pulse correlation of each independent component, and correct the parasitic correlation; According to the corrected parasitic correlation of each independent component, combined with the independent weight of each independent component in each time period, reconstruct the first current data.

[0056] The main source of the parasitic effect is the charging and discharging of the parasitic capacitance, and the change of the pulse width will affect the charging degree and discharging process of the parasitic capacitance, thereby affecting the leakage current. As the pulse width gradually increases, the parasitic effect in the protected circuit also increases. Therefore, in the first current data, the proportion of the current fluctuation caused by the parasitic effect is higher; According to the independent component decomposition results of the first current data in the corresponding time period of each pulse, combined with the width of each pulse current, obtain the pulse correlation of each independent component: Take the independent weight of each independent component in each time period as the numerator; Take the time length of each time period as the denominator; Extract the dispersion degree of the corresponding fractions of each independent component in all time periods, and take the result of the negative correlation mapping of the dispersion degree as the pulse correlation of each independent component. In this embodiment, the standard deviation is used to calculate the dispersion degree of multiple variables. Denote the dispersion degree of each independent component as A, then the formula for the pulse correlation of each independent component is: ; where is a preset parameter to prevent the denominator from being 0, and the value in this embodiment is 1.

[0057] It should be understood that when the change of the independent weight of each independent component in all time periods is similar to the change of the pulse width, the dispersion degree of the corresponding ratio is relatively low, and at this time, the pulse correlation of the corresponding independent component is higher.

[0058] Further, according to the pulse correlation of each independent component, the parasitic correlation is corrected: the normalized value of the pulse correlation of each independent component and the positive fusion result of the parasitic correlation are used as the corrected parasitic correlation of each independent component. In this embodiment, the normalization method uses the exponential normalization method, and the positive fusion between multiple variables uses the multiplication method.

[0059] According to the corrected parasitic correlation of each independent component, for the independent components with strong correlation, the influence of the parasitic effect of the protected device on them is relatively large. Therefore, these components need to be suppressed and then the leakage current is reconstructed.

[0060] The opposite number of the corrected parasitic correlation of each independent component is normalized, and the product of the obtained normalized result and the independent weight of each independent component in each time period is used as the corrected weight of each independent component in each time period. In this embodiment, the normalization method uses the exponential normalization method.

[0061] According to the corrected weight of each independent component in each time period, the first current data is reconstructed. Denote the corrected first current data in the r-th time period as: the linear combination of each independent component and its corrected weight in the r-th time period; the specific formula form is: ; where represents the corrected weight of the i-th independent component in the r-th time period; n represents the number of independent components; represents the i-th independent component; represents the reconstructed first current data.

[0062] Among them, the schematic diagram for obtaining the reconstruction of the first current data is as Figure 2 shown.

[0063] So far, the first current data after removing the interference of the parasitic effect is obtained.

[0064] Step Five: Detect the ESD device based on the deviation between the reconstructed first current data and the static current data.

[0065] Measure the static current under the normal working voltage (such as 3.3V). According to the first current data after removing the interference of the parasitic effect, observe whether the maximum deviation from the static current is higher than the preset value of 1 μA. When it is higher than the preset value, the ESD device cannot effectively protect the circuit in the corresponding pulse discharge, and the ESD device may be damaged or has entered the deteriorated state.

[0066] Through the systematic test method, it can be ensured that the ESD protection device can work effectively under different ESD events, avoiding problems of protection failure or mis-triggering.

[0067] Based on the same inventive concept as the above method, an embodiment of the present application further provides a detection system for an ESD device used in the FAN-OUT process, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods for detecting an ESD device used in the FAN-OUT process.

[0068] In summary, the present application obtains current data under a transmission line pulse and decomposes the current data between the ESD protector and the protected device, fully considering the influence of the parasitic effect, effectively suppressing the components greatly affected by the parasitic effect, thereby improving the accuracy of detection. Specifically, the present application decomposes the current data between the ESD protector and the protected device, calculates the parasitic correlation between each independent component and the operating current of the protected circuit, and combines the strength characteristics of the parasitic effect under different pulse widths to correct the parasitic correlation, and then reconstructs more realistic current data. Based on the optimized current data, the present application can more accurately evaluate the key parameters of the ESD protection device and ensure the reliability of the detection result.

[0069] In addition, by testing the protection ability and tolerance of the ESD device under different pulse current conditions, the present application provides a systematic and standardized test process, which helps to reduce the measurement error and misjudgment risk caused by the parasitic effect and improve the ESD protection optimization efficiency in the product design stage. Finally, this method not only improves the performance evaluation accuracy of the ESD protection device, but also reduces the later repair and rework costs, enhances the overall reliability and market competitiveness of the product, and has important engineering application value for the ESD protection detection in the FAN-OUT process.

[0070] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a portion of code that contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions noted in the blocks may occur in an order different from that noted in the accompanying drawings. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functionality involved. In the description corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in an order different from that disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functionality involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0071] For those skilled in the art, it is obvious that the present application is not limited to the details of the above-described exemplary embodiments, and can be implemented in other specific forms without departing from the basic characteristics of the present application. Therefore, from any point of view, the above-described embodiments of the present application should be regarded as exemplary and non-limiting; modifications to the technical solutions described in the foregoing embodiments, or equivalent replacements of some of the technical features, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included within the protection scope of the present application.

Claims

1. Detection method for ESD device used in FAN-OUT process, characterized in that, The method includes the following steps: Obtain the first current data between the ESD protector and the device to be protected; obtain the second current data between the protector and the power supply; Based on the number of data in the longest time period corresponding to the pulse, perform interpolation processing on the first current data and the second current data for each time period respectively; perform independent component analysis on the first current data for all time periods corresponding to the pulse to obtain a preset number of independent components of the first current data and their independent weights in each time period; Analyze the falling characteristics of the second current data for each time period to obtain the falling part of the second current data for each time period, and analyze the similarity characteristics between each independent component of the first current data and the falling part of the second current data for each time period to obtain the parasitic correlation of each independent component; According to the independent component decomposition result of the first current data in the time period corresponding to each pulse, combined with the width of the current of each pulse, obtain the pulse correlation of each independent component, and correct the parasitic correlation; according to the corrected parasitic correlation of each independent component combined with the independent weight of each independent component in each time period, reconstruct the first current data; Based on the deviation between the reconstructed first current data and the current data under the preset voltage, detect the ESD device.

2. The detection method of the ESD device for the FAN-OUT process according to claim 1, characterized in that, The specific process of obtaining the falling part of the second current data for each time period is as follows: Denote the sequence composed of the interpolated second current data for each time period as the second current data sequence; obtain the second-order difference sequence of the second current data sequence; Based on the element distribution characteristics in the second-order difference sequence, obtain the segmentation moment, divide the second current data in the corresponding time period into two parts, and denote the later segmentation part as the falling part.

3. The detection method of the ESD device for the FAN-OUT process according to claim 2, wherein, The segmentation moment is specifically the moment corresponding to the smallest minimum value point in the second-order difference sequence.

4. The detection method of the ESD device for the FAN-OUT process according to claim 1, wherein, The specific process of obtaining the parasitic correlation of each independent component is as follows: Obtain the subsequence of each independent component of the first current data after the segmentation moment corresponding to each time period; obtain the similarity measure between the subsequence obtained for each independent component in each time period and the falling part of the second current data in the corresponding time period; take the mean of the similarity measures obtained for each independent component in all time periods as the parasitic correlation of each independent component.

5. The detection method of the ESD device for the FAN-OUT process according to claim 1, wherein, The specific process of obtaining the pulse correlation of each independent component is as follows: Take the independent weight of each independent component in each time period as the numerator; take the time length of each time period as the denominator; extract the discreteness of the corresponding fractions of each independent component in all time periods, and take the result of the negative correlation mapping of the discreteness as the pulse correlation of each independent component.

6. The detection method of the ESD device for the FAN-OUT process according to claim 1, wherein, The specific process of correcting the parasitic correlation is as follows: take the normalized value of the pulse correlation of each independent component and the positive fusion result of the parasitic correlation as the corrected parasitic correlation of each independent component.

7. The detection method of the ESD device for the FAN-OUT process according to claim 1, wherein The specific process of reconstructing the first current data is as follows: Based on the corrected parasitic correlation of each independent component, correct the independent weight of each independent component in each time period to obtain the corrected weight; Use the linear combination of each independent component and its corrected weight in each time period as the first current data reconstructed for each time period.

8. The detection method of the ESD device for the FAN-OUT process according to claim 7, characterized in that, The correction process of the independent weight of each independent weight in each time period is as follows: normalize the opposite number of the corrected parasitic correlation of each independent component, and use the product of the obtained normalization result and the independent weight of each independent component in each time period as the corrected weight of each independent component in each time period.

9. The detection method of the ESD device for the FAN-OUT process according to claim 1, characterized in that, The detection of the ESD device is specifically as follows: Record the current data under the preset voltage as the static current; Calculate the maximum deviation between the reconstructed first current data and the static current; when the maximum deviation is greater than the preset value, it is determined that there is a problem with the ESD device; otherwise, it is determined that there is no problem with the ESD device.

10. A detection system for an ESD device used in a FAN-OUT process, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-9.

Images