Computer mainboard test system and test method

Abstract

The invention discloses a computer mainboard testing system and a computer mainboard testing method, and relates to the technical field of mainboard testing, the system obtains a perturbation characteristic data set by performing high-precision synchronous sampling on mainboard electrical nodes, and obtains a perturbation characteristic data set by constructing a micro-variation amplitude disturbance index var, a time sequence signal texture energy index str and a high-frequency signal disturbance center offset index shf; and a mainboard perturbation health degree index MHI is calculated, and quantitative health assessment of the electrical texture structure is realized. When abnormal fluctuation of the texture structure is detected, thermosensitive aging analysis is further executed, a dynamic aging thermosensitive index DAF is constructed, and the stability of the mainboard under the thermal state operation condition is evaluated. The system can complete mainboard health grade division based on multi-stage health assessment results, and generate a detection health report which can be accessed to a factory management system, thereby realizing hidden fault identification and thermal driving type early aging risk monitoring of the computer mainboard, and remarkably improving test accuracy and factory screening efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of motherboard testing, and in particular to a computer motherboard testing system and a testing method. Background Art

[0002] With the increasing demand for performance and stability in modern computing devices, the quality and reliability of computer motherboards, as the core carrier, directly determine the upper limit of overall system performance. During the manufacturing process, efficient and accurate functional testing of motherboards has become a critical step in ensuring product quality. Currently, traditional motherboard test systems primarily focus on functional verification and parameter limit testing, making it difficult to identify potential hidden failure characteristics of motherboards under high-frequency operation and thermal loads. Especially in the context of increasingly complex multi-phase power supply designs and high-speed bus transmission structures, dynamic aging symptoms such as micro-perturbations, electrical texture changes, and signal structure migration under thermal conditions exhibited during motherboard operation have become significant factors affecting their long-term stability. Therefore, precise quantification and systematic evaluation of the electrical texture structural health and thermally induced aging response characteristics of computer motherboards have become a key direction for upgrading motherboard testing systems.

[0003] The current mainstream testing methods mostly rely on functional-level detection or static threshold judgment, and lack a detailed characterization of the motherboard's micro-perturbation behavior, electrical texture energy distribution, and high-frequency disturbance characteristics, resulting in insufficient perception of early latent failure phenomena. In addition, although some systems have introduced temperature rise testing methods, they mostly focus on thermal protection mechanisms or thermal failure point judgment, and fail to effectively couple the dynamic disturbance behavior under thermal-sensitive conditions with the texture stability characteristics during operation. This makes it impossible to achieve linkage modeling and joint identification of the variation trend of electrical texture health signals and thermal anomalies in actual evaluations, and there are significant blind spots in the test coverage. In particular, the ability to identify thermal-sensitive early-aging motherboards before they leave the factory is significantly insufficient, affecting the consistency and reliability control of the final product. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the present invention provides a computer motherboard testing system and testing method, which solve the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A computer motherboard testing system includes a sampling channel configuration module, a data monitoring module, a perturbation feature analysis module, a thermal response analysis module and an intelligent judgment module; The sampling channel configuration module is used to connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same start signal source through the unified Trigger bus; The data monitoring module is used to monitor the signal data of the computer motherboard in real time based on the sampling device, and perform preprocessing to obtain the perturbation characteristic data group; The perturbation feature analysis module is used to perform summary calculations based on the perturbation feature data group to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. The module then performs comprehensive calculations to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. The thermal response analysis module is used to execute the motherboard thermal aging analysis instruction when the electrical texture structure health assessment shows abnormal fluctuations, calculate the dynamic aging thermal index DAF based on the motherboard perturbation health index MHI, and perform the motherboard thermal stability assessment; The intelligent judgment module is used to classify health levels based on the assessment results and generate health report data, which is then connected to the factory management system in real time through a dedicated API interface protocol.

[0006] Preferably, the sampling channel configuration module includes a test node design unit and a synchronization control unit; The test node design unit is used to identify and calibrate the motherboard test point pads, debug interface, exposed pads and chip resistor connection ends based on the motherboard schematic and wiring diagram, and uses probe cards, spring contacts and debug header interfaces for contact-type non-welding access. The synchronization control unit is used to connect all sampling channels to the same start signal source through a unified trigger bus to establish a synchronous sampling structure. The start signal trigger edge is uniformly issued by the test system control main control module, and the FPGA logic judgment module is used to implement the trigger source arbitration and buffer logic to adjust the timing coordination and out-of-step protection mechanism between the samplers; All sampled data are packaged into data frames according to a unified time base, and the signals of each channel are fully aligned within each frame.

[0007] Preferably, the data monitoring module includes a data monitoring unit and a data processing unit; The data acquisition unit is used to monitor the signal data of the computer motherboard in real time based on the sampling equipment connected to the motherboard test point; Sampling equipment includes high-resolution ADCs, non-contact high-frequency current probes, differential probes, oscilloscopes, clock jitter analyzers, and spectrum analyzers; Directly collect the signal edge jitter frequency fs through the clock jitter analyzer; Directly collect the center frequency zf of the main peak of the spectrum through the spectrum analyzer; A high-resolution ADC is used to collect the voltage signal v in real time from the power supply path between the multiphase voltage regulator module on the computer motherboard and the power input pins of the load chip. The acquisition time window T is set. The high-resolution ADC is used to extract the peak value of the voltage signal v within each acquisition time window T. The difference between the peak values of each two adjacent voltage signal v segments is then calculated using the differential method to obtain the peak-to-peak voltage difference ∆vp. Use a non-contact high-frequency current probe with a bandwidth greater than 100MHz to connect to the mainboard power output channel, set the sampling frequency to 500MS / s, connect the probe signal output to a bandpass filter, set the passband frequency to 2MHz to 20MHz, isolate the high-frequency disturbance component ih in the current, and then perform a fast Fourier transform (FFT) on the high-frequency disturbance component ih of each window segment to obtain the spectrum distribution Ih(f) of each window segment {f1, f2}. Calculate the disturbance energy density P of each window segment {f1, f2} based on the spectrum distribution Ih(f). Then, after normalizing the disturbance energy density under all windows, calculate the average value to obtain the current disturbance density ih, specifically: , , where W represents the total number of acquisition window segments, and df represents the small frequency calculus amount; Use a differential probe to connect to the mainboard bus test point and use an oscilloscope to collect the original signal s(t). Set the oscilloscope bandwidth to greater than 5 GHz and the sampling rate to greater than 20 GS / s. Then divide the signal into H frames according to the window, with the time width of each frame being Δt. Calculate the power gl of each frame and use the statistical method to calculate the signal power texture density sd. Specifically, , , where Δgl represents the difference in power between adjacent frames, represents the mean value of all frame powers; The data processing unit is used to perform denoising, outlier processing and dimensionless processing on the monitored signal data to obtain a perturbation characteristic data group; The perturbation characteristic data set includes the voltage peak-to-peak difference ∆vp, the current perturbation density ih, the signal power texture density sd, the signal edge jitter frequency fs and the center frequency of the spectrum main peak zf.

[0008] Preferably, the perturbation feature analysis module includes a perturbation feature analysis unit and a perturbation health analysis unit; The perturbation feature analysis unit is used to perform summary calculations based on the perturbation feature data group to obtain the slight change amplitude perturbation index var, the time series signal texture energy index str, and the high-frequency signal perturbation center offset index shf. The specific formula is as follows: ; ; ; Where N represents the number of sampling tests on the mainboard electrical nodes, ∆vp i Represents the peak-to-peak voltage difference of the i-th node, ih i represents the current disturbance density of the i-th node, K represents the number of timing test channels, sd j represents the signal power texture density of the jth channel, fs jrepresents the jitter frequency of the jth channel, log represents the logarithmic function, M represents the number of high-frequency signal channels, zf m (t) represents the center frequency of the main peak of the spectrum of the mth channel at time t, ∆t represents the sampling interval, zf m (t-∆t) represents the center frequency of the main peak of the spectrum of the mth channel in the previous time period.

[0009] Preferably, the perturbation health analysis unit includes an electrical texture health analysis unit and an electrical texture health assessment unit; The electrical texture health analysis unit is used to perform comprehensive calculations based on the micro-variation amplitude disturbance index var, the timing signal texture energy index str, and the high-frequency signal disturbance center offset index shf to obtain the motherboard micro-perturbation health index MHI. The specific formula is as follows: ; In the formula, log represents the logarithmic function, Represents a small positive constant, with a value of 1×10 −6 .

[0010] Preferably, the electrical texture health assessment unit is used to sort all historical motherboard perturbation health indexes MHI from small to large, and use the percentile statistics method to set the historical motherboard perturbation health index MHI at 10% as the first electrical texture structure health threshold A, and the historical motherboard perturbation health index MHI at 30% as the second electrical texture structure health threshold B, and then compare them with the motherboard perturbation health index MHI obtained in real time, and perform electrical texture structure health assessment based on the comparison results. The specific assessment scheme is as follows; When the motherboard micro-perturbation health index MHI is less than the first electrical texture structure health threshold A, it means that the motherboard signal texture structure is stable and the motherboard has no hidden problems; When the first electrical texture structure health threshold A ≤ motherboard perturbation health index MHI < the second electrical texture structure health threshold B, it indicates that there is abnormal fluctuation in the motherboard signal texture structure, and the motherboard thermal aging analysis instruction is executed at this time; When the motherboard micro-perturbation health index MHI ≥ the second electrical texture structure health threshold B, it indicates that the motherboard signal texture structure is offset and the motherboard shows signs of aging. At this time, the motherboard is marked as a risky motherboard and relevant technicians are notified to isolate and disassemble it.

[0011] Preferably, the thermal response analysis module includes a thermal aging analysis unit and a thermal stability evaluation unit; The thermal aging analysis unit is used to execute the mainboard thermal aging analysis instruction when the electrical texture structure health assessment shows that there is abnormal fluctuation; The motherboard thermal aging analysis command is executed to calculate the dynamic aging thermal index DAF based on the motherboard micro-perturbation health index MHI. The specific formula is as follows: ; Where, MHI id Indicates the motherboard perturbation health index under idle standby conditions, ∆T cpu Indicates the temperature change of the motherboard CPU, which is obtained by subtracting the CPU temperature in standby state from the real-time CPU temperature.

[0012] Preferably, the thermal stability evaluation unit is used to calculate the average of all historical dynamic aging thermal sensitivity indexes DAF according to a statistical method, preset a thermal induced aging threshold C based on the average, and compare it with the dynamic aging thermal sensitivity index DAF obtained in real time, and perform a motherboard thermal stability evaluation based on the comparison result. The specific evaluation scheme is as follows; When the dynamic aging thermal sensitivity index DAF is less than the thermal induced aging threshold C, it means that the motherboard is thermally stable and in a healthy state, and is marked as shipped normally; When the dynamic aging thermal sensitivity index DAF ≥ the thermal induced aging threshold C, it indicates that the thermal state of the motherboard is unstable and the motherboard is abnormal. In this case, the motherboard is marked as a maintenance motherboard.

[0013] Preferably, the intelligent judgment module is used to classify the health level of the computer motherboard according to the results of the electrical texture structure health assessment and the motherboard thermal stability assessment; A motherboard that is assessed to have a stable signal texture structure is classified as a first healthy motherboard; Classify a motherboard that is assessed to be thermally stable as the second healthiest motherboard; Classify a motherboard that is assessed to be thermally unstable as a third healthier motherboard; Classifying a motherboard that is assessed to have a shifted signal texture structure as a fourth healthy motherboard; Based on the health level, the system generates a unique identification structure inspection health report data for the motherboard, converts the generated health report into PDF format for data packaging, and then connects with the factory management system in real time through a dedicated API interface protocol. When the electrical texture structure health assessment and the motherboard thermal stability assessment detect signal texture structure deviation and thermal instability, an alert is immediately triggered and uploaded to the cloud platform and remote server for archiving.

[0014] A computer motherboard testing method comprises the following steps: S1. Connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same startup signal source through the unified Trigger bus; S2. monitoring the signal data of the computer mainboard in real time using a sampling device, and performing pre-processing to obtain a perturbation characteristic data set; S3. Perform summary calculation based on the perturbation feature data set to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. Then, perform comprehensive calculation to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. S4. When the electrical texture structure health assessment indicates abnormal fluctuations, execute the motherboard thermal aging analysis instruction, calculate the dynamic aging thermal index DAF based on the motherboard micro-perturbation health index MHI, and perform the motherboard thermal stability assessment; S5. Health levels are divided according to the assessment results, and health report data is generated, which is then connected to the factory management system in real time through a dedicated API interface protocol.

[0015] The present invention provides a computer motherboard testing system and testing method. It has the following beneficial effects: (1) The system automatically analyzes the motherboard schematic and wiring diagram through the sampling channel configuration module, and accurately calibrates and identifies the test point pads, debug interfaces, exposed pads, and chip resistor connection terminals on the motherboard with the help of the test node design unit. It also uses probe cards, spring contacts, and debug header interfaces for physical access through contact-type non-welding methods, solving the problem of traditional welding test methods being difficult to quickly deploy and repeatedly access. At the same time, the system's built-in synchronization control unit connects all sampling channels through a unified Trigger bus, accesses the same start signal source, and performs trigger source arbitration and buffer logic adjustment through the FPGA logic judgment module to ensure the timing of test startup is coordinated and consistent, preventing the samplers from losing step due to master synchronization anomalies. All sampling data are packaged into data frames according to a unified time base, ensuring that the signal data of different test channels are completely aligned on the timeline, providing high-precision data support for cross-channel texture feature construction, thereby significantly improving the sampling data quality, test deployment efficiency, and the reliability of timing analysis.

[0016] (2) The system data monitoring module and the perturbation feature analysis module together constitute a high-resolution and high-time efficiency electrical perturbation modeling platform. The data monitoring module is equipped with a high-resolution ADC, a non-contact high-frequency current probe, a differential probe, an oscilloscope, a clock jitter analyzer, and a spectrum analyzer, which respectively realize the real-time acquisition of various electrical features such as voltage signals, current perturbations, bus timing signals, and high-frequency noise, and construct a perturbation feature data set. The perturbation feature analysis module calculates the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf in sequence, and then forms the motherboard perturbation health index MHI index that uniformly expresses the health status of the motherboard signal texture through log function adjustment and parameter normalization. The MHI index is then used to evaluate the electrical texture structure health with the first electrical texture structure health threshold A and the second electrical texture structure health threshold B, effectively quantifying the motherboard's potential signal stability problems, transient power supply degradation characteristics, and high-frequency perturbation performance, significantly improving the recognition rate of soft failures and early abnormalities.

[0017] (3) The system’s thermal response analysis module and intelligent judgment module realize multi-level evaluation of the motherboard’s thermal response capability and overall health level. The thermal aging analysis unit dynamically calculates the thermal aging index DAF based on the motherboard’s micro-perturbation health index MHI evaluation results, and uses the motherboard’s micro-perturbation health index MHI to calculate the thermal aging index DAF when the CPU temperature changes ∆T. cpu The system identifies thermally driven aging phenomena by analyzing the amplitude of fluctuations under load. The dynamic aging thermal sensitivity index (DAF) is calculated to further reflect the motherboard's declining signal stability under load. The thermal stability assessment unit then constructs a thermally induced aging threshold (C) based on the statistical mean of all historical DAF values. This threshold is then compared with the real-time DAF value to accurately determine whether thermal stability meets factory requirements. When the motherboard's micro-perturbation health index (MHI) is abnormal and the DAF exceeds the corresponding threshold, the system automatically classifies the motherboard's health level through an intelligent judgment module. A uniquely identified health report data is automatically packaged in PDF format and pushed to the factory management system via a dedicated API. This enables full-process integration of test data with dispatch, warehousing, and repair processes. Furthermore, offset anomaly information is uploaded to a cloud platform for archiving in real time, enabling rapid isolation, tracking, and recall of abnormal motherboards. This multi-module collaborative mechanism not only improves test response speed and result utilization, but also enhances automation and management precision throughout the motherboard testing process through sophisticated dynamic assessment and real-time reporting. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a computer motherboard testing system according to the present invention; Figure 2 This is a schematic diagram of the steps of a computer motherboard testing method of the present invention; Figure 3 The present invention is a schematic diagram of the operating principle of a computer motherboard testing system. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] Example 1 See also Figure 1 and Figure 3 The present invention provides a computer motherboard testing system. To achieve the above purpose, the present invention is implemented through the following technical solutions: including a sampling channel configuration module, a data monitoring module, a perturbation feature analysis module, a thermal response analysis module and an intelligent judgment module; The sampling channel configuration module is used to connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same start signal source through the unified Trigger bus; The data monitoring module is used to monitor the signal data of the computer motherboard in real time based on the sampling device, and perform preprocessing to obtain the perturbation characteristic data group; The perturbation feature analysis module is used to perform summary calculations based on the perturbation feature data group to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. The module then performs comprehensive calculations to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. The thermal response analysis module is used to execute the motherboard thermal aging analysis instruction when the electrical texture structure health assessment shows abnormal fluctuations, calculate the dynamic aging thermal index DAF based on the motherboard perturbation health index MHI, and perform the motherboard thermal stability assessment; The intelligent judgment module is used to classify health levels based on the assessment results and generate health report data, which is then connected to the factory management system in real time through a dedicated API interface protocol.

[0021] In this embodiment, the sampling channel configuration module automatically identifies the locations of pads, exposed pads, and debug ports based on the motherboard schematic and wiring diagram, and implements test access through non-welding methods such as probe cards and spring contacts, significantly improving the speed and reusability of test deployment. At the same time, a unified trigger bus connects to the same startup signal source, and in conjunction with the FPGA logic judgment structure, a synchronous sampling mechanism for all channels is established, effectively avoiding data mismatch problems caused by asynchronous startup signals. This module achieves time base alignment of multi-channel test signals, providing a high-precision data foundation for subsequent texture structure analysis, significantly improving the scalability and engineering feasibility of the test system. The data monitoring module, based on multi-source sampling equipment such as high-resolution ADCs, differential probes, jitter analyzers, and spectrum analyzers, collects and processes signal data from the computer motherboard to form a perturbation feature data set. Feature fusion constructs the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf, which are then integrated into the motherboard perturbation health index (MHI) to comprehensively characterize the stability of the motherboard's electrical texture structure. By modeling and analyzing fine-grained electrical disturbances, this module not only enables early detection of soft anomalies such as power regulation instability and excessive signal jitter, but also overcomes the subjective limitations of traditional methods that rely on visual waveform inspection and empirical judgment, significantly improving the sensitivity and quantification of motherboard health testing. The thermal response analysis module and intelligent judgment module enable comprehensive, intelligent decision-making regarding the motherboard's thermal stability and health level. When the motherboard's Micro-Perturbation Health Index (MHI) indicates abnormal fluctuations in the texture structure, the thermal response analysis module immediately activates thermally driven aging analysis logic. Based on the dynamic changes in the MHI, the Dynamic Aging Thermal Index (DAF) is calculated. Combined with real-time temperature differences, it identifies trends in signal response degradation under high-temperature conditions. This approach addresses a gap in existing systems' thermal anomaly detection capabilities and effectively determines whether a motherboard has thermally induced stability risks. The intelligent judgment module then integrates the electrical health and thermal stability assessment results to assign a health level and automatically flag abnormal motherboards. This generates health report data, which is then connected to the factory management system via a dedicated API, enabling real-time, closed-loop integration of test data with the production dispatch system. Compared with the traditional method of manual recording and local report output, this module significantly improves the utilization rate of test results, the speed of abnormal response and the level of manufacturing process automation, thereby bringing significant systematic progress in product yield improvement, test efficiency optimization, and quality closed-loop assurance.

[0022] Example 2 Please refer to Figure 1 ,Specifically: the sampling channel configuration module includes a test node design unit and a ,synchronization control unit; The test node design unit is used to identify and calibrate the motherboard test point pads, debug interface, exposed pads, and chip resistor connection ends based on the motherboard schematic and wiring diagram. It uses probe cards, spring contacts, and debug header interfaces for contact-type, non-welding access. If the motherboard design reserves a debug port, a numbered cable adapter is used for direct connection. The synchronization control unit is used to connect all sampling channels to the same start signal source through a unified trigger bus to establish a synchronous sampling structure. The start signal trigger edge is uniformly issued by the test system control main control module, and the FPGA logic judgment module is used to implement the trigger source arbitration and buffer logic to adjust the timing coordination and out-of-step protection mechanism between the samplers; All sampling data are packaged into data frames according to a unified time base, and the channel signals in each frame are completely aligned to meet the needs of cross-channel texture construction.

[0023] In this embodiment, the sampling channel configuration module realizes the precise identification and non-welding rapid access of the motherboard test points through the collaborative construction of the test node design unit and the synchronization control unit, supports multiple contact methods such as probe cards, spring contacts and DebugHeader interfaces, and improves the flexibility and efficiency of test deployment; on this basis, all sampling channels are connected to the same start signal source through a unified Trigger bus, and the trigger source is arbitrated and the buffer logic is adjusted in combination with the FPGA logic judgment module, thereby constructing a synchronous sampling structure with timing coordination and out-of-step protection capabilities, so that all channel acquisition data is output in frame form under a unified time base, realizing cross-channel signal alignment. This implementation method not only improves the data consistency and availability in high-density motherboard test scenarios, but also overcomes the test error problems caused by complex access methods and poor signal synchronization in traditional tests, thereby achieving higher-precision support for subsequent electrical texture structure analysis, and significantly improving the overall test reliability, deployment convenience and analysis accuracy of the system.

[0024] Example 3 Please refer to Figure 1 ,Specifically: the data monitoring module includes a data monitoring unit and a data ,processing unit; The data acquisition unit is used to monitor the signal data of the computer motherboard in real time based on the sampling equipment connected to the motherboard test point; Sampling equipment includes high-resolution ADCs, non-contact high-frequency current probes, differential probes, oscilloscopes, clock jitter analyzers, and spectrum analyzers; The signal edge jitter frequency fs is directly collected by the clock jitter analyzer to suppress the energy increase under high jitter. The spectrum analyzer directly collects the center frequency zf of the main peak of the spectrum, which indicates the stability of the center frequency of the main peak position of the spectrum. It is used to capture the frequency drift phenomenon in the frequency domain that cannot be measured by the time domain logic analyzer. A high-resolution ADC is used to acquire the voltage signal v in real time from the power supply path between the multiphase voltage regulator module on the computer motherboard and the power input pins of the load chip. An acquisition time window T is set. The high-resolution ADC extracts the peak value of the voltage signal v within each acquisition time window T. The difference between the peak values of two adjacent voltage signal v segments is then calculated using the differential method to obtain the peak-to-peak voltage difference ∆vp. This indicates the presence of chronic hysteresis and parasitic oscillation in the response of the capacitor and inductor at a specific point. This is used to capture the changing trends of the instantaneous power supply regulation performance and reveal the degradation trend of the voltage regulation loop. A non-contact high-frequency current probe with a bandwidth greater than 100 MHz is connected to the mainboard power output channel. The sampling frequency is set to 500 MS / s. The probe signal output is connected to a bandpass filter with a passband frequency set to 2 MHz to 20 MHz. The high-frequency disturbance component ih in the current is isolated. A fast Fourier transform (FFT) is then performed on the high-frequency disturbance component ih in each window segment to obtain the spectrum distribution Ih(f) of each window segment {f1, f2}. The disturbance energy density P of each window segment {f1, f2} is calculated based on the spectrum distribution Ih(f). After normalizing the disturbance energy density under all windows, the mean is calculated to obtain the current disturbance density ih, which represents the density of current fluctuations in each power supply branch within the sampling window. It reflects the consistency of the load transient response and is used to detect subtle soft load anomalies and non-ideal power supply behavior problems. Specifically: , , where W represents the total number of acquisition window segments, and df represents the small frequency calculus amount; Use a differential probe to connect to the mainboard bus test point and use an oscilloscope to collect the original signal s(t). Set the oscilloscope bandwidth to greater than 5 GHz and the sampling frequency to greater than 20 GS / s. Then divide the signal into H frames according to the window, with the time width of each frame being Δt. Calculate the power gl of each frame and use a statistical method to calculate the signal power texture density sd, which represents the energy density of the signal texture and reflects whether there are concentrated energy clusters or noise spurious peaks in the signal spectrum. It is used to quantify the stability characteristics of the timing signal. The more concentrated the energy clusters are, the higher the signal consistency is. Specifically: , , where Δgl represents the difference in power between adjacent frames, represents the mean value of all frame powers; The data processing unit is used to perform denoising, outlier processing and dimensionless processing on the monitored signal data to obtain a perturbation characteristic data group; Denoising uses wavelet transform technology to decompose signals at different frequency scales to eliminate the influence of noise in the data. Outlier processing uses the interquartile range method to detect and process outliers in the signal data. Dimensionless processing uses the Max-Min maximum minimization method to eliminate the dimensional influence of the signal data. The perturbation characteristic data set includes the voltage peak-to-peak difference ∆vp, the current perturbation density ih, the signal power texture density sd, the signal edge jitter frequency fs and the center frequency of the spectrum main peak zf.

[0025] In this embodiment, the data monitoring module is collaboratively composed of a data monitoring unit and a data processing unit. First, by integrating a high-resolution ADC, a non-contact high-frequency current probe, a differential probe, a clock jitter analyzer, a spectrum analyzer, and a high-bandwidth oscilloscope, high-precision, multi-dimensional real-time acquisition of electrical signals at multiple points on the computer motherboard is achieved. Five key perturbation parameters, including the voltage peak-to-peak difference ∆vp, the current perturbation density ih, the signal power texture density sd, the signal edge jitter frequency fs, and the spectrum main peak center frequency zf, are extracted in sequence to reflect the voltage regulation response, power supply fluctuation density, timing consistency, clock stability, and frequency center drift characteristics, respectively, to form a perturbation feature data set. Subsequently, frequency domain decomposition and denoising are performed using wavelet transform, outliers are eliminated using the interquartile range method, and dimensionality effects are eliminated using the maximum and minimum normalization method to improve data consistency and analysis reliability. The implementation of this module not only achieves in-depth extraction of fine-grained and structural features of the motherboard's electrical signals, but also breaks through the limitations of traditional voltage / current average measurement methods that cannot identify soft failure phenomena such as dynamic fluctuations, frequency offsets, and timing anomalies. It significantly improves the sensitivity, accuracy, and diagnostic depth of motherboard dynamic behavior monitoring, providing a high-quality data foundation for subsequent health modeling and risk assessment, thereby optimizing the overall detection system's intelligence level and process defect screening capabilities.

[0026] Example 4 Please refer to Figure 1 ,Specifically: the perturbation feature analysis module includes a perturbation feature analysis unit and a perturbation health ,analysis unit; The perturbation feature analysis unit is used to perform summary calculations based on the perturbation feature data group to obtain the slight change amplitude disturbance index var, the time series signal texture energy index str and the high-frequency signal disturbance center offset index shf; The slight change amplitude disturbance index var is used to measure the intensity of slight voltage and current disturbances at all test points during the sampling period. The specific formula is as follows: ; Where N represents the number of sampling tests on the mainboard electrical nodes, ∆vp i Represents the peak-to-peak voltage difference of the i-th node, ih i represents the current disturbance density of the i-th node; The timing signal texture energy index str is used to measure the disturbance energy characteristics of multiple groups of logic timing signals. The specific formula is as follows: ; Where K represents the number of timing test channels, sdj represents the signal power texture density of the jth channel, fs j represents the jitter frequency of the jth channel, log represents the logarithmic function, This item controls the effect of high jitter on energy values, preventing high-frequency but low-energy jitter from being amplified and enhancing the ability to express stable features. The high-frequency signal disturbance center shift index (shf) measures the center shift trend of the high-frequency disturbance signal and reflects the center frequency shift caused by capacitor and clock source aging. ; Where M represents the number of high-frequency signal channels, zf m (t) represents the center frequency of the main peak of the spectrum of the mth channel at time t, ∆t represents the sampling interval, zf m (t-∆t) represents the center frequency of the main peak of the spectrum of the mth channel in the previous time period.

[0027] The perturbation health analysis unit includes an electrical texture health analysis unit and an electrical texture health assessment unit; The electrical texture health analysis unit is used to perform comprehensive calculations based on the micro-variation amplitude disturbance index var, the timing signal texture energy index str, and the high-frequency signal disturbance center offset index shf to obtain the motherboard micro-perturbation health index MHI. The specific formula is as follows: ; In the formula, log represents the logarithmic function, which is used to control the smoothness of the value range, avoid the explosive amplification of a single factor, and control the consistency of the data scale. Represents a small positive constant, with a value of 1×10 −6 , used to prevent the denominator from being zero and protect the continuity and robustness of the formula structure, It represents the countervailing relationship between the texture energy index of the time series signal and the center shift index of the high-frequency signal disturbance.

[0028] The electrical texture health assessment unit is used to sort all historical motherboard perturbation health indices (MHIs) from small to large, and use the percentile statistics method to set the historical motherboard perturbation health index (MHI) at 10% as the first electrical texture structure health threshold A, and the historical motherboard perturbation health index (MHI) at 30% as the second electrical texture structure health threshold B. These are then compared with the real-time motherboard perturbation health index (MHI), and an electrical texture structure health assessment is performed based on the comparison results. The specific assessment plan is as follows; When the motherboard micro-perturbation health index MHI is less than the first electrical texture structure health threshold A, it means that the motherboard signal texture structure is stable and the motherboard has no hidden problems; When the first electrical texture structure health threshold A ≤ motherboard perturbation health index MHI < the second electrical texture structure health threshold B, it indicates that there is abnormal fluctuation in the motherboard signal texture structure, and the motherboard thermal aging analysis instruction is executed at this time; When the motherboard micro-perturbation health index MHI ≥ the second electrical texture structure health threshold B, it indicates that the motherboard signal texture structure is offset and the motherboard shows signs of aging. At this time, the motherboard is marked as a risky motherboard and relevant technicians are notified to isolate and disassemble it.

[0029] In this embodiment, based on the collected perturbation feature data set, the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf are calculated. These three factors are then integrated within the perturbation health analysis unit through logarithmic functions and constant adjustments to calculate the motherboard perturbation health index (MHI). The system then uses percentile statistics to establish a first electrical texture health threshold A and a second electrical texture health threshold B based on all historical motherboard perturbation health index (MHI) values, performing a graded assessment of motherboard health. This enables quantitative modeling and dynamic assessment of signal texture stability, timing integrity, and spectral offset trends. This implementation effectively addresses the technical shortcomings of traditional test systems, such as the inability to perceive perturbation evolution trends and the lack of structural quantitative diagnosis. This approach provides greater sensitivity and recognition of early-stage hidden problems, thermally driven aging, and signal frequency drift anomalies, significantly improving the accuracy of motherboard fault identification, the intelligence of the detection process, and the stability and process optimization capabilities of overall system testing.

[0030] Example 5 Please refer to Figure 1 ,Specifically: the thermal response analysis module includes a thermal aging analysis unit and a thermal stability evaluation unit; The thermal aging analysis unit is used to execute the mainboard thermal aging analysis instruction when the electrical texture structure health assessment shows that there is abnormal fluctuation; Executing the motherboard thermal aging analysis command is used to calculate the dynamic aging thermal index (DAF) based on the motherboard's micro-perturbation health index (MHI). This analyzes whether the motherboard exhibits signal perturbations under thermal loads and determines whether there are thermally driven premature aging issues. The specific formula is as follows: ; Where, MHI id Indicates the motherboard perturbation health index under idle standby conditions, ∆T cpu Indicates the temperature change of the motherboard CPU, which is obtained by subtracting the CPU temperature in standby state from the real-time CPU temperature.

[0031] The thermal stability assessment unit is used to calculate the mean of all historical dynamic aging thermal sensitivity indices (DAF) using a statistical method. Based on this mean, a preset thermal induced aging threshold (C) is set. This is then compared with the real-time dynamic aging thermal sensitivity indices (DAF). Based on this comparison, the motherboard thermal stability is assessed. The specific assessment scheme is as follows: When the dynamic aging thermal sensitivity index DAF is less than the thermal induced aging threshold C, it means that the motherboard is thermally stable and in a healthy state, and is marked as shipped normally; When the dynamic aging thermal sensitivity index DAF ≥ thermal induced aging threshold C, it indicates that the motherboard is thermally unstable and abnormal. At this time, the motherboard is marked as a maintenance motherboard and relevant technicians are notified to perform abnormal maintenance.

[0032] In this embodiment, when the electrical texture structure health index MHI of the motherboard fluctuates abnormally, the thermal aging analysis instruction is automatically triggered. id Real-time CPU temperature change ∆T cpu , calculates the dynamic aging thermal index DAF, accurately assesses the motherboard's signal disturbances under thermal load conditions, and thus identifies whether there are heat-driven early aging issues. Subsequently, the thermal-induced aging threshold C is set by statistically averaging historical dynamic aging thermal index DAF data, and compared with the real-time dynamic aging thermal index DAF to determine and label the motherboard's thermal stability level. This module effectively addresses the lack of thermal stability assessment in existing testing systems, and can identify high-temperature-induced signal structure degradation before the motherboard exhibits overt functional failures. This improves the proactiveness and scientific nature of anomaly identification, ultimately achieving accurate screening and automatic isolation of risky motherboards, significantly improving the long-term reliability and operational stability of factory-produced motherboards.

[0033] Example 6 Please refer to Figure 1 ,Specifically: the intelligent judgment module is used to classify the health level of the ,computer motherboard based on the results of the electrical texture structure health ,assessment and the motherboard thermal stability assessment; A motherboard that is assessed to have a stable signal texture structure is classified as a first healthy motherboard; Classify a motherboard that is assessed to be thermally stable as the second healthiest motherboard; Classify a motherboard that is assessed to be thermally unstable as a third healthier motherboard; Classifying a motherboard that is assessed to have a shifted signal texture structure as a fourth healthy motherboard; Based on the health level, the system generates a unique identification structure inspection health report data for the motherboard, converts the generated health report into PDF format for data packaging, and then connects with the factory management system in real time through a dedicated API interface protocol to achieve automatic test dispatch and warehousing linkage, improve test data utilization and process flow efficiency, and immediately trigger a prompt when signal texture structure offset and thermal instability are detected in the electrical texture structure health assessment and motherboard thermal stability assessment and upload to the cloud platform and remote server for archiving.

[0034] In this embodiment, each motherboard under test is assigned a multi-dimensional health grade based on the micro-perturbation health index (MHI) derived from the electrical texture health assessment and the dynamic aging thermal sensitivity index (DAF) output by the thermal response analysis module. Motherboards with stable signal texture structures are labeled as healthy (1st), thermally stable (2nd), thermally unstable but electrically normal (3rd), and those with deviated electrical texture structures (4th). Each grade corresponds to a uniquely identified health report data volume. The report is automatically packaged into a PDF format and connected to the factory management system via a dedicated API protocol. This enables automatic classification of motherboard health status, digital archiving of test results, real-time notification and upload of abnormal motherboards, and automatic dispatching and warehousing of normal motherboards. This module implements a closed-loop health assessment from electrical to thermal status, connecting the entire process from testing to production management. This significantly improves the automation level of motherboard testing, the speed of abnormality response, data traceability, and production line scheduling efficiency. It also addresses the problems of manual grading lag, abnormality tracking difficulties, and information silos associated with traditional testing methods.

[0035] Example 7 Please refer to Figure 2 , a computer motherboard testing method, comprising the following steps: S1. Connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same startup signal source through the unified Trigger bus; S2. monitoring the signal data of the computer mainboard in real time using a sampling device, and performing pre-processing to obtain a perturbation characteristic data set; S3. Perform summary calculation based on the perturbation feature data set to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. Then, perform comprehensive calculation to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. S4. When the electrical texture structure health assessment indicates abnormal fluctuations, execute the motherboard thermal aging analysis instruction, calculate the dynamic aging thermal index DAF based on the motherboard micro-perturbation health index MHI, and perform the motherboard thermal stability assessment; S5. Health levels are divided according to the assessment results, and health report data is generated, which is then connected to the factory management system in real time through a dedicated API interface protocol.

[0036] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A computer motherboard testing system, characterized by: It includes sampling channel configuration module, data monitoring module, perturbation feature analysis module, thermal response analysis module and intelligent judgment module; The sampling channel configuration module is used to connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same start signal source through the unified Trigger bus; The data monitoring module is used to monitor the signal data of the computer motherboard in real time based on the sampling device, and perform preprocessing to obtain the perturbation characteristic data group; The perturbation feature analysis module is used to perform summary calculations based on the perturbation feature data group to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. The module then performs comprehensive calculations to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. The thermal response analysis module is used to execute the motherboard thermal aging analysis instruction when the electrical texture structure health assessment shows abnormal fluctuations, calculate the dynamic aging thermal index DAF based on the motherboard perturbation health index MHI, and perform the motherboard thermal stability assessment; The intelligent judgment module is used to classify health levels based on the assessment results and generate health report data, which is then connected to the factory management system in real time through a dedicated API interface protocol.

2. A computer motherboard testing system according to claim 1, characterized in that: The sampling channel configuration module includes a test node design unit and a synchronization control unit; The test node design unit is used to identify and calibrate the motherboard test point pads, debug interface, exposed pads and chip resistor connection ends based on the motherboard schematic and wiring diagram, and uses probe cards, spring contacts and debug header interfaces for contact-type non-welding access. The synchronization control unit is used to connect all sampling channels to the same start signal source through a unified trigger bus to establish a synchronous sampling structure. The start signal trigger edge is uniformly issued by the test system control main control module, and the FPGA logic judgment module is used to implement the trigger source arbitration and buffer logic to adjust the timing coordination and out-of-step protection mechanism between the samplers; All sampled data are packaged into data frames according to a unified time base, and the signals of each channel are fully aligned within each frame.

3. A computer motherboard testing system according to claim 2, characterized in that: The data monitoring module includes a data monitoring unit and a data processing unit; The data acquisition unit is used to monitor the signal data of the computer motherboard in real time based on the sampling equipment connected to the motherboard test point; Sampling equipment includes high-resolution ADCs, non-contact high-frequency current probes, differential probes, oscilloscopes, clock jitter analyzers, and spectrum analyzers; Directly collect the signal edge jitter frequency fs through the clock jitter analyzer; Directly collect the center frequency zf of the main peak of the spectrum through the spectrum analyzer; A high-resolution ADC is used to collect the voltage signal v in real time from the power supply path between the multiphase voltage regulator module on the computer motherboard and the power input pins of the load chip. The acquisition time window T is set. The high-resolution ADC is used to extract the peak value of the voltage signal v within each acquisition time window T. The difference between the peak values of each two adjacent voltage signal v segments is then calculated using the differential method to obtain the peak-to-peak voltage difference ∆vp. Use a non-contact high-frequency current probe with a bandwidth greater than 100MHz to connect to the mainboard power output channel, set the sampling frequency to 500MS / s, connect the probe signal output to a bandpass filter, set the passband frequency to 2MHz to 20MHz, isolate the high-frequency disturbance component ih in the current, and then perform a fast Fourier transform (FFT) on the high-frequency disturbance component ih of each window segment to obtain the spectrum distribution Ih(f) of each window segment {f1, f2}. Calculate the disturbance energy density P of each window segment {f1, f2} based on the spectrum distribution Ih(f). Then, after normalizing the disturbance energy density under all windows, calculate the average value to obtain the current disturbance density ih, specifically: , , where W represents the total number of acquisition windows, and df represents the small frequency calculus; Use a differential probe to connect to the mainboard bus test point and use an oscilloscope to collect the original signal s(t). Set the oscilloscope bandwidth to greater than 5 GHz and the sampling rate to greater than 20 GS / s. Then divide the signal into H frames according to the window, with the time width of each frame being Δt. Calculate the power gl of each frame and use the statistical method to calculate the signal power texture density sd. Specifically, , , where Δgl represents the difference in power between adjacent frames, represents the mean value of all frame powers; The data processing unit is used to perform denoising, outlier processing and dimensionless processing on the monitored signal data to obtain a perturbation characteristic data group; The perturbation characteristic data set includes the voltage peak-to-peak difference ∆vp, the current perturbation density ih, the signal power texture density sd, the signal edge jitter frequency fs and the center frequency of the spectrum main peak zf.

4. A computer motherboard testing system according to claim 3, characterized in that: The perturbation feature analysis module includes a perturbation feature analysis unit and a perturbation health analysis unit; The perturbation feature analysis unit is used to perform summary calculations based on the perturbation feature data group to obtain the slight change amplitude perturbation index var, the time series signal texture energy index str, and the high-frequency signal perturbation center offset index shf. The specific formula is as follows: ； ； ； Where N represents the number of sampling tests on the mainboard electrical nodes, ∆vp i Represents the peak-to-peak voltage difference of the i-th node, ih i represents the current disturbance density of the i-th node, K represents the number of timing test channels, sd j represents the signal power texture density of the jth channel, fs j represents the jitter frequency of the jth channel, log represents the logarithmic function, M represents the number of high-frequency signal channels, zf m (t) represents the center frequency of the main peak of the spectrum of the mth channel at time t, ∆t represents the sampling interval, zf m (t-∆t) represents the center frequency of the main peak of the spectrum of the mth channel in the previous time period.

5. A computer motherboard testing system according to claim 4, characterized in that: The perturbation health analysis unit includes an electrical texture health analysis unit and an electrical texture health assessment unit; The electrical texture health analysis unit is used to perform comprehensive calculations based on the micro-variation amplitude disturbance index var, the timing signal texture energy index str, and the high-frequency signal disturbance center offset index shf to obtain the motherboard micro-perturbation health index MHI. The specific formula is as follows: ； In the formula, log represents the logarithmic function, Represents a small positive constant, with a value of 1×10 −6 .

6. A computer motherboard testing system according to claim 5, characterized in that: The electrical texture health assessment unit is used to sort all historical motherboard perturbation health indices (MHIs) from small to large, and use the percentile statistics method to set the historical motherboard perturbation health index (MHI) at 10% as the first electrical texture structure health threshold A, and the historical motherboard perturbation health index (MHI) at 30% as the second electrical texture structure health threshold B. These are then compared with the real-time motherboard perturbation health index (MHI), and an electrical texture structure health assessment is performed based on the comparison results. The specific assessment plan is as follows; When the motherboard micro-perturbation health index MHI is less than the first electrical texture structure health threshold A, it means that the motherboard signal texture structure is stable and the motherboard has no hidden problems; When the first electrical texture structure health threshold A ≤ motherboard perturbation health index MHI < the second electrical texture structure health threshold B, it indicates that there is abnormal fluctuation in the motherboard signal texture structure, and the motherboard thermal aging analysis instruction is executed at this time; When the motherboard micro-perturbation health index MHI ≥ the second electrical texture structure health threshold B, it indicates that the motherboard signal texture structure is offset and the motherboard shows signs of aging. At this time, the motherboard is marked as a risky motherboard and relevant technicians are notified to isolate and disassemble it.

7. A computer motherboard testing system according to claim 6, characterized in that: The thermal response analysis module includes a thermal aging analysis unit and a thermal stability evaluation unit; The thermal aging analysis unit is used to execute the mainboard thermal aging analysis instruction when the electrical texture structure health assessment shows that there is abnormal fluctuation; The motherboard thermal aging analysis command is executed to calculate the dynamic aging thermal index DAF based on the motherboard micro-perturbation health index MHI. The specific formula is as follows: ； Where, MHI id Indicates the motherboard perturbation health index under idle standby conditions, ∆T cpu Indicates the temperature change of the motherboard CPU, which is obtained by subtracting the CPU temperature in standby state from the real-time CPU temperature.

8. A computer motherboard testing system according to claim 7, characterized in that: The thermal stability assessment unit is used to calculate the mean of all historical dynamic aging thermal sensitivity indices (DAF) using a statistical method. Based on this mean, a preset thermal induced aging threshold (C) is set. This is then compared with the real-time dynamic aging thermal sensitivity indices (DAF). Based on this comparison, the motherboard thermal stability is assessed. The specific assessment scheme is as follows: When the dynamic aging thermal sensitivity index DAF is less than the thermal induced aging threshold C, it means that the motherboard is thermally stable and in a healthy state, and is marked as shipped normally; When the dynamic aging thermal sensitivity index DAF ≥ the thermal induced aging threshold C, it indicates that the thermal state of the motherboard is unstable and the motherboard is abnormal. In this case, the motherboard is marked as a maintenance motherboard.

9. A computer motherboard testing system according to claim 1, characterized in that: The intelligent judgment module is used to classify the health level of the computer motherboard based on the results of the electrical texture structure health assessment and the motherboard thermal stability assessment; A motherboard that is assessed to have a stable signal texture structure is classified as a first healthy motherboard; Classify a motherboard that is assessed to be thermally stable as the second healthiest motherboard; Classify a motherboard that is assessed to be thermally unstable as the third healthiest motherboard; Classifying a motherboard that is assessed to have a shifted signal texture structure as a fourth healthy motherboard; Based on the health level, the system generates a unique identification structure inspection health report data for the motherboard, converts the generated health report into PDF format for data packaging, and then connects with the factory management system in real time through a dedicated API interface protocol. When the electrical texture structure health assessment and the motherboard thermal stability assessment detect signal texture structure deviation and thermal instability, an alert is immediately triggered and uploaded to the cloud platform and remote server for archiving.

10. A computer motherboard testing method, applied to a computer motherboard testing system according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1. Connect the motherboard test points according to the motherboard schematic and wiring diagram, and connect to the same startup signal source through the unified Trigger bus; S2. monitoring the signal data of the computer mainboard in real time using a sampling device, and performing pre-processing to obtain a perturbation characteristic data set; S3. Perform summary calculation based on the perturbation feature data set to obtain the micro-variation amplitude perturbation index var, the timing signal texture energy index str, and the high-frequency signal perturbation center offset index shf. Then, perform comprehensive calculation to obtain the motherboard perturbation health index MHI for electrical texture structure health assessment. S4. When the electrical texture structure health assessment indicates abnormal fluctuations, execute the motherboard thermal aging analysis instruction, calculate the dynamic aging thermal index DAF based on the motherboard perturbation health index MHI, and perform a motherboard thermal stability assessment; S5. Health levels are divided according to the assessment results, and health report data is generated, which is then connected to the factory management system in real time through a dedicated API interface protocol.

Images