Mainboard testing method and system

Abstract

The invention provides a mainboard testing method and system, and relates to the technical field of mainboard testing methods.The mainboard testing method comprises the steps that firstly, an impedance compensation matrix is generated through pre-calibration sweep frequency signals, full-band dynamic calibration is conducted on probe impedance characteristics, and signal attenuation and phase deviation caused by impedance mismatch between a testing probe and a mainboard interface are eliminated; secondly, based on multi-channel frequency domain filtering and linear combination, separating frequency domain sensitive frequency bands and time domain fluctuation characteristics of test parameters, calculating parameter weights by fusing frequency domain energy and time domain variance, avoiding the limitation of single domain analysis, and enhancing the detection sensitivity of the hidden defects of the mainboard; according to the test parameter weight vector and the probe weight mapping matrix, the test parameter weight vector is converted into a test probe weight vector, finally, the parameter importance is quantified as a basis for probe excitation intensity and sampling rate adjustment, and the beneficial effect that test resources are distributed according to needs is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of motherboard testing methods, and particularly to a motherboard testing method and its system. Background Art

[0002] Traditional motherboard testing methods generally adopt fixed-frequency excitation signals and static probe parameter configurations, making it difficult to meet the testing requirements in high-frequency and multi-parameter scenarios. Due to the impedance characteristics of the probe and the motherboard interface changing dynamically with frequency, the manual calibration or narrowband compensation methods in the prior art are prone to cause measurement distortion of wideband signals. Moreover, when analyzing the time-domain and frequency-domain characteristics separately, the weight allocation of key parameters depends on empirical thresholds, and it is impossible to accurately quantify the influence of sensitive frequency bands and timing fluctuations on the test results. In addition, the excitation intensity and sampling rate of the test probes are usually fixed, making it difficult to dynamically optimize resource allocation according to real-time signal characteristics, resulting in low testing efficiency. Especially in the testing of high-density motherboards, problems such as missed detections or over-detections are likely to occur. Therefore, there is an urgent need for a motherboard testing method that can adaptively calibrate broadband impedance, fuse time-frequency domain feature analysis, and dynamically regulate testing resources to improve the testing accuracy and efficiency of complex motherboards.

[0003] Therefore, it is necessary to provide a motherboard testing method and its system to solve the above technical problems. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a motherboard testing method and its system, achieving the beneficial effects of adaptive broadband impedance calibration, fusing time-frequency domain feature analysis, and dynamically regulating testing resources.

[0005] The present invention provides a motherboard testing method, and the testing method includes the following steps:

[0006] S1: Apply a pre-calibrated swept-frequency signal to the motherboard to be tested, obtain the impedance data of each test probe at different frequency points, and generate an impedance compensation matrix;

[0007] S2: Apply a test excitation signal to the motherboard to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency-domain conversion on all the original signals of the test probes, and then perform frequency-domain calibration using the impedance compensation matrix to obtain the frequency-domain signals of each calibrated test probe;

[0008] S3: Perform multi-channel frequency-domain filtering and linear combination processing on the frequency-domain signals of each calibrated test probe to obtain the frequency-domain signals of each test parameter, extract frequency-domain features, convert the frequency-domain signals of each test parameter into time-domain signals, extract time-domain features, and calculate the test parameter weight vector based on the time-domain features and frequency-domain features;

[0009] S4: According to the preset probe weight mapping matrix, convert the test parameter weight vector into a test probe weight vector;

[0010] S5: Adjust the excitation intensity and sampling rate of each test probe according to a preset adjustment rule based on the test probe weight vector.

[0011] Preferably, the frequency range of the pre-calibrated sweep signal covers 1.5 times the fundamental frequency of the main board.

[0012] Preferably, the calculation formula of the impedance compensation matrix is:

[0013]

[0014] where K is the impedance compensation matrix, k n,f is the impedance compensation coefficient of test probe n at frequency point f, is a set of complex numbers, N is the total number of test probes, F is the total number of frequency points, Z ref (f) is the preset reference impedance of the test probe at frequency point f, |Z ref (f)| is the amplitude of the preset reference impedance of the test probe at frequency point f, Z n (f) is the actual impedance of test probe n at frequency point f, |Z n (f)| is the amplitude of the actual impedance of test probe n at frequency point f, ∠Z n (f) is the phase of the actual impedance of test probe m at frequency point f, ∠Z ref (f) is the phase of the preset reference impedance of the test probe at frequency point f, V n (f) is the complex voltage of test probe n at frequency point f, I n (f) is the complex current of test probe m at frequency point f, and j is the imaginary unit.

[0015] Preferably, in step S2, the formula for frequency domain calibration using the impedance compensation matrix after frequency domain conversion of all the original signals of the test probes is:

[0016]

[0017] where S n (f) is the original signal of test probe n at frequency point f, is the calibrated signal of test probe n at frequency point f.

[0018] Preferably, the generation steps of the test parameter weight vector are:

[0019] Perform multi-channel frequency domain filtering and linear combination processing on the frequency domain signals of each calibrated test probe to obtain the frequency domain signals of each test parameter;

[0020] Extract the frequency domain energy of the key frequency points of the frequency domain signals of each test parameter within the preset sensitive frequency band as the frequency domain feature;

[0021] Perform an inverse Fourier transform on the frequency-domain signals of each test parameter to obtain the time-domain signals of each test parameter and calculate the time-domain variance as the time-domain feature;

[0022] Based on the frequency-domain features and time-domain features, calculate the importance coefficients of each test parameter and normalize the importance coefficients of all obtained test parameters to obtain the test parameter weight vector.

[0023] Preferably, the calculation formula for the importance coefficient is:

[0024]

[0025] where, R p is the importance coefficient of test parameter p, is the time-domain variance of test parameter p, E p (f s ) is the frequency-domain energy of the test parameter at the key frequency point f s , and λ is the time-domain weight coefficient.

[0026] Preferably, the calculation formula for the test parameter weight vector is:

[0027]

[0028] μ = [μ1, μ2, ……, μ P

[0029] where, μ p is the test parameter weight of test parameter p, and μ is the test parameter weight vector.

[0030] Preferably, the calculation formula for the probe weight mapping matrix is:

[0031] M = [m p,n P×N

[0032]

[0033] w = [ω1, ω2, ……, ω N

[0034] where, M is the probe weight mapping matrix, ω n is the test probe weight of test probe n, P is the total number of test parameters, m p,n is the weight of the dependence of test parameter p on test probe n, μ p is the test parameter weight of test parameter p, and w is the test probe weight vector.

[0035] Preferably, the calculation formula for the weight of the dependence of test parameter p on test probe n is:​​​

[0036]

[0037] wherein, m p,n is the weight of the dependence of the test parameter p on the test probe n, and D p (f) is the sensitivity of the test parameter p at the frequency point f.

[0038] The present invention also provides a main board test system, which is applied to a main board test method. The test system includes:

[0039] A probe impedance calibration module, configured to apply a pre-calibrated sweep signal to the main board to be tested, obtain impedance data of each test probe at different frequency points, and generate an impedance compensation matrix;

[0040] A frequency domain calibration module, configured to apply a test excitation signal to the main board to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency domain conversion on all the original signals of the test probes, and then perform frequency domain calibration by using the impedance compensation matrix to obtain the frequency domain signals of each calibrated test probe;

[0041] A multi-domain feature fusion module, configured to perform multi-channel frequency domain filtering and linear combination processing on the frequency domain signals of each calibrated test probe, obtain the frequency domain signals of each test parameter and extract frequency domain features, convert the frequency domain signals of each test parameter into time domain signals and extract time domain features, and calculate a test parameter weight vector based on the time domain features and the frequency domain features;

[0042] A weight mapping conversion module, configured to convert the test parameter weight vector into a test probe weight vector according to a preset probe weight mapping matrix;

[0043] An excitation regulation module, configured to adjust the excitation intensity and sampling rate of each test probe based on the test probe weight vector through a preset adjustment rule.

[0044] Compared with the related art, a main board test method and system provided by the present invention have the following beneficial effects:

[0045] The present invention first generates an impedance compensation matrix covering 1.5 times the fundamental frequency through pre-calibrated swept-frequency signals, dynamically calibrates the impedance characteristics of the probes across the entire frequency band, eliminates signal attenuation and phase shift caused by impedance mismatch between the test probes and the main board interface, and significantly improves the measurement accuracy of high-frequency signals. Then, based on multi-channel frequency-domain filtering and linear combination, it separates the frequency-domain sensitive bands and time-domain fluctuation characteristics of the test parameters, calculates the parameter weights by fusing the frequency-domain energy and time-domain variance, avoids the limitations of single-domain analysis, and enhances the detection sensitivity to hidden defects on the main board. Next, according to the test parameter weight vector and the probe weight mapping matrix, the test parameter weight vector is transformed into a test probe weight vector. Finally, the parameter importance is quantified as the basis for adjusting the probe excitation intensity and sampling rate, realizing the on-demand allocation of test resources, increasing the excitation power for high-weight probes to enhance the signal-to-noise ratio, and simultaneously increasing the sampling rate to capture transient anomalies, while reducing the resource occupancy for low-weight probes, thereby achieving the beneficial effects of adaptive broadband impedance calibration, fusing time-frequency domain feature analysis, and dynamically regulating test resources. Description of the Drawings

[0046] Figure 1 is a flowchart of a main board test method of the present invention;

[0047] Figure 2 is a module structure diagram of a main board test system of the present invention. Detailed Embodiments

[0048] The present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only for explaining the present invention, rather than limiting the present invention. Additionally, it should be noted that for the sake of description, only parts related to the present invention are shown in the drawings rather than all structures. Furthermore, the embodiments in the present invention and the features in the embodiments can be combined with each other without conflict.

[0049] It should also be noted that for the sake of description, only parts related to the present invention are shown in the drawings rather than all content. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as being processed sequentially, many of the operations can be performed in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operations are completed, but it can also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0050] Embodiment 1

[0051] A main board test method, in the specific implementation process, as Figure 1As shown, it shows a flowchart of a motherboard testing method provided by the present invention, including:

[0052] Step S1: Apply a pre-calibrated sweep signal to the motherboard to be tested, obtain impedance data of each test probe at different frequency points, and generate an impedance compensation matrix.

[0053] Specifically, the frequency range of the pre-calibrated sweep signal covers 1.5 times the fundamental frequency of the motherboard.

[0054] Specifically, the calculation formula for the impedance compensation matrix is:

[0055]

[0056] Where K is the impedance compensation matrix, k n,f is the impedance compensation coefficient of test probe n at frequency point f, is a set of complex numbers, N is the total number of test probes, F is the total number of frequency points, Z ref (f) is the preset reference impedance of the test probe at frequency point f, |Z ref (f)| is the amplitude of the preset reference impedance of the test probe at frequency point f, Z n (f) is the actual impedance of test probe n at frequency point f, |Z n (f)| is the amplitude of the actual impedance of test probe n at frequency point f, ∠Z n (f) is the phase of the actual impedance of test probe n at frequency point f, ∠Z ref (f) is the phase of the preset reference impedance of the test probe at frequency point f, V n (f) is the complex voltage of test probe n at frequency point f, I n (f) is the complex current of test probe n at frequency point f, and j is the imaginary unit.

[0057] In the specific implementation process, impedance deviation caused by probe individual differences, transmission line losses, and environmental interference needs to be eliminated before a batch of motherboard tests. Therefore, it is necessary to obtain impedance data of a test probe at different frequency points through a pre-calibration sweep signal and generate an impedance compensation matrix to correct the test signal in subsequent tests. First, a pre-calibrated sweep signal is applied to the motherboard, the fundamental frequency in the motherboard specification is read, and the sweep range is automatically generated. The lower limit is the larger of 10% of the fundamental frequency or 10 MHz, and the upper limit is 1.5 times the fundamental frequency. A direct digital frequency synthesizer is used to generate a discrete sweep sequence, and through a vector signal generator, the sweep signal is synchronously output to all test probes. The complex voltage and complex current of test probe n at frequency point f are synchronously collected, and the actual impedance of test probe n is calculated. The impedance compensation coefficient of test probe n at frequency point f is calculated by combining the reference impedance of each test probe at each frequency point. Exemplarily, the nominal impedance value is extracted from the interface specification document as the reference impedance. The reference impedance represents the ideal impedance characteristics of the test system under no interference conditions, and its core function is to provide a reference value for actual measurement. Among them, V n (f) and I n (f) are obtained by synchronously collecting the time-domain signals of the probes and performing I / Q demodulation. Their real and imaginary parts respectively represent the amplitude components of the signal in the orthogonal coordinate system. This process directly reflects the actual coupling state and transmission path loss between the probe and the motherboard test point. The generation of the compensation coefficient is achieved by multiplying the amplitude ratio of the reference impedance to the actual impedance by the phase difference compensation term. Its physical meaning is to eliminate the impedance deviation caused by probe individual differences, transmission line losses, and environmental interference through amplitude normalization and phase alignment operations, so that the response characteristics of all probes at any frequency point approach the ideal reference model. The finally constructed impedance compensation matrix stores the compensation coefficients of all probe and frequency point combinations in complex form.

[0058] Step S2: Apply a test excitation signal to the motherboard to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency-domain conversion on all the original signals of the test probes, and then perform frequency-domain calibration using the impedance compensation matrix to obtain the frequency-domain signals of each calibrated test probe.

[0059] Specifically, in step S2, the formula for performing frequency-domain calibration on all the original signals of the test probes using the impedance compensation matrix after frequency-domain conversion is:

[0060]

[0061] where S n (f) is the original signal of test probe n at frequency point f, is the calibrated signal of test probe n at frequency point f.

[0062] In the specific implementation process, after obtaining the impedance compensation matrix, the main board is tested. Exemplarily, first, an excitation signal matching the main board interface protocol is generated by a multi-channel arbitrary waveform generator, and is synchronously applied to the main board test points via the gold finger test probe array. The FPGA trigger module is used to insert a delay of 5 ns ± 20 ps after the rising edge of the excitation signal to avoid switching noise. Subsequently, the original time-domain signals of each probe are synchronously captured at a sampling rate of 12.5 GS / s. By converting the original test probe signals collected within a preset sliding time window to the frequency domain, a complex frequency-domain spectrum is generated. Then, the pre-stored impedance compensation matrix is called to dynamically index the corresponding compensation coefficients according to the current excitation frequency point, and a point-by-point complex multiplication operation is performed on the frequency-domain spectrum of each probe. Finally, the frequency-domain signals of each calibrated test probe are generated.

[0063] Step S3: Perform multi-channel frequency-domain filtering and linear combination processing on the frequency-domain signals of each calibrated test probe to obtain the frequency-domain signals of each test parameter and extract the frequency-domain features. Convert the frequency-domain signals of each test parameter to time-domain signals and extract the time-domain features. Calculate the test parameter weight vector based on the time-domain features and frequency-domain features.

[0064] In the specific implementation process, it is necessary to separate the signals of each test parameter from the signals in the test probe, then extract the features from the signals of the test parameter, quantify the features and calculate the test parameter weight vector. The test parameter weight vector is used as the basis for calculating the weight vector of the subsequent test probe.

[0065] Specifically, the generation steps of the test parameter weight vector are as follows:

[0066] Perform multi-channel frequency-domain filtering and linear combination processing on the frequency-domain signals of each calibrated test probe to obtain the frequency-domain signals of each test parameter;

[0067] Extract the frequency-domain energy at the key frequency points within the preset sensitive frequency band of the frequency-domain signals of each test parameter as the frequency-domain feature;

[0068] Perform inverse Fourier transform on the frequency-domain signals of each test parameter to obtain the time-domain signals of each test parameter and calculate the time-domain variance as the time-domain feature;

[0069] Based on the frequency-domain features and time-domain features, calculate the importance coefficients of each test parameter and normalize the obtained importance coefficients of all test parameters to obtain the test parameter weight vector.

[0070] In the specific implementation process, first, a predefined sensitive frequency band filter bank is configured according to the type of test parameters. The frequency domain signals of each calibrated test probe are linearly combined through a preset probe space weight matrix to generate the frequency domain signals of each parameter, and the frequency domain energy of preset key frequency points is extracted. Subsequently, the frequency domain signals are converted into time domain waveforms through inverse fast Fourier transform, and the time domain variance is calculated. Through the importance coefficient calculation formula, the frequency domain energy of the key frequency points and the time domain variance are fused to obtain the importance coefficient of the test parameter, where the time domain weight coefficient λ is optimized and determined through cross-validation.

[0071] Specifically, the calculation formula for the importance coefficient is:

[0072]

[0073] where R p is the importance coefficient of the test parameter p, is the time domain variance of the test parameter p, and E p (f s ) is the frequency domain energy of the test parameter at the key frequency point f s , and λ is the time domain weight coefficient.

[0074] Specifically, the calculation formula for the test parameter weight vector is:

[0075]

[0076] μ = [μ1, μ2, ……, μ P

[0077] where μ p is the test parameter weight of the test parameter p, and μ is the test parameter weight vector.

[0078] In the specific implementation process, the importance coefficient of the test parameter is obtained through the importance coefficient calculation formula, and then the obtained importance coefficient of the test parameter is transformed into a test parameter weight vector through normalization formula processing. The test parameter weight vector is used as the calculation basis for the subsequent test probe weight vector.

[0079] Step S4: According to the preset probe weight mapping matrix, transform the test parameter weight vector into a test probe weight vector.

[0080] Specifically, the calculation formula for the probe weight mapping matrix is:

[0081] M = [m p,n P×N

[0082]

[0083] w = [ω1, ω2, ……, ω​​N

[0084] Among them, M is the probe weight mapping matrix, and ω n is the test probe weight of test probe n, P is the total number of test parameters, and m p,n is the weight of the dependence of test parameter p on test probe n, and μ p is the test parameter weight of test parameter p, and w is the test probe weight vector.

[0085] Specifically, the calculation formula for the weight of the dependence of test parameter p on test probe n is:

[0086]

[0087] Among them, m p,n is the weight of the dependence of test parameter p on test probe n, and D p (f) is the sensitivity of test parameter p at frequency point f.

[0088] In the specific implementation process, first, based on the preset probe weight mapping matrix, the test parameter weight vector is transformed into a test probe weight vector. The core process of this step is to calculate the elements of the probe weight mapping matrix through frequency-domain sensitivity analysis. The sensitivity of probe n at frequency point f is calibrated through experiments, and the test probe weight of test probe n is calculated through the formula. The test probe weight is used as the basis for resource allocation adjustment during the test process.

[0089] Step S5: Based on the test probe weight vector, adjust the excitation intensity and sampling rate of each test probe through the preset adjustment rules.

[0090] In the specific implementation process, by way of example, the probe weights are divided into three levels. For high-weight probes, which are related to key signal links, the excitation intensity and sampling accuracy need to be maximized. For medium-weight probes, which are secondary signal or power supply monitoring probes, the default configuration is maintained. For low-weight probes, which are non-critical probes, such as auxiliary power supply ground monitoring, an energy-saving mode is enabled. In addition, by way of example, the excitation intensity adjustment formula is Among them, A n is the adjusted excitation intensity of test probe n, A base is the reference excitation intensity, which is determined according to the interface protocol of the motherboard under test. α is the attenuation coefficient, which is calibrated through experiments to control the sensitivity of the excitation intensity changing with the weight. The sampling rate adjustment formula is Among them, F n is the sampling rate of test probe n, β is the scaling factor, which is set according to the hardware performance, ω n is the probe weight of test probe n, f max is the maximum frequency of the current test frequency band, which is defined by the protocol, f base ​Taking the reference frequency, usually the interface base frequency, adjusting the excitation intensity and sampling rate through the test probe weights, converting the probe weights into operable hardware parameters, achieving adaptive broadband impedance calibration, fusing time-frequency domain feature analysis, and dynamically regulating test resources.

[0091] The working principle of a motherboard testing method provided by the present invention is as follows:

[0092] First of all, the present invention uses a pre-calibrated swept-frequency signal covering 1.5 times the fundamental frequency of the motherboard to measure the complex impedance of each test probe and construct an impedance compensation matrix, which includes amplitude and phase compensation coefficients for subsequent signal calibration. Then, after applying the test excitation signal, the original time-domain signal is transformed to the frequency domain by Fourier transform, and the signal distortion caused by the probe impedance difference is eliminated through matrix compensation. Subsequently, the calibrated multi-channel frequency-domain signals are filtered and linearly combined to separate the independent frequency-domain components of each test parameter, synchronously extract the energy features of the key frequency bands, namely the frequency-domain features and the time-domain variance, i.e., the time-domain features, fuse the two types of features to calculate the parameter importance coefficient, and generate a normalized test parameter weight vector. Then, through a preset probe weight mapping matrix, the test parameter weight vector is transformed into a test probe weight vector. Finally, according to the weights, the excitation intensity and sampling rate of each probe are dynamically adjusted to achieve the intelligent allocation of test resources to the key signal paths. This method improves the signal fidelity through impedance compensation, combines the time-frequency domain feature fusion to evaluate the parameter importance, and establishes a weight conduction mechanism between the test parameters and the test probes, forming a closed-loop optimization system, significantly improving the test accuracy and efficiency, and at the same time realizing the reasonable allocation of test resources.

[0093] Embodiment 2

[0094] The present invention also provides a motherboard testing system applied to a motherboard testing method. In the specific implementation process, as Figure 2 shown, it shows a schematic diagram of the module structure of a motherboard testing system provided by the present invention, including:

[0095] A probe impedance calibration module 100, configured to apply a pre-calibrated swept-frequency signal to the motherboard to be tested, obtain the impedance data of each test probe at different frequency points, and generate an impedance compensation matrix;

[0096] A frequency-domain calibration module 200, configured to apply a test excitation signal to the motherboard to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency-domain conversion on all the original signals of the test probes, and then perform frequency-domain calibration using the impedance compensation matrix to obtain the frequency-domain signals of each calibrated test probe;

[0097] The multi-domain feature fusion module 300 is used to perform multi-channel frequency-domain filtering and linear combination processing on the frequency-domain signals of the calibrated test probes, obtain the frequency-domain signals of each test parameter and extract frequency-domain features, convert the frequency-domain signals of each test parameter into time-domain signals and extract time-domain features, and calculate the test parameter weight vector based on the time-domain features and frequency-domain features;

[0098] The weight mapping conversion module 400 is used to convert the test parameter weight vector into a test probe weight vector according to the preset probe weight mapping matrix;

[0099] The excitation regulation module 500 is used to adjust the excitation intensity and sampling rate of each test probe based on the test probe weight vector through the preset adjustment rules.

[0100] The working principle of a motherboard test system provided by the present invention is as follows:

[0101] A motherboard test system of the present invention is based on a closed-loop optimization architecture with multi-module collaboration. Its technical principle realizes the optimal allocation of test resources through progressive processing of impedance calibration, signal calibration, feature fusion, and dynamic regulation. The specific system consists of five core modules. The probe impedance calibration module 100 applies a pre-calibration sweep signal covering 1.5 times the fundamental frequency, and calculates the impedance compensation matrix of each probe based on the complex impedance formula to complete the calibration of the amplitude and phase compensation coefficients of the impedance differences between the probes; the frequency-domain calibration module 200 performs frequency-domain calibration on multi-channel signals using the impedance compensation matrix after Fourier-transforming the original signal during the test stage to eliminate the measurement deviation introduced by the probe characteristics; the multi-domain feature fusion module 300 performs frequency-domain filtering and linear combination on the calibrated signals to separate test parameters, synchronously extracts the energy of the key frequency band, i.e., the frequency-domain feature, and the time-domain variance, i.e., the time-domain feature, fuses the frequency-domain feature and the time-domain feature to calculate the parameter importance coefficient and generate a normalized weight vector; the weight mapping conversion module 400 converts the test parameter weight vector into a test probe weight vector based on the preset probe weight mapping matrix. Finally, the excitation regulation module 500 dynamically adjusts the excitation intensity and sampling rate of each probe according to the weight vector, forming an adaptive closed-loop from signal acquisition to feature analysis to weight calculation and finally resource reallocation, realizing the intelligent focusing of test resources on the key signal path, effectively improving the test accuracy and efficiency, and realizing the reasonable allocation of test resources at the same time.

[0102] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.

[0103] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program, and this program can be stored in a computer-readable storage medium. The storage medium includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM), or other optical disc memories, magnetic disk memories, tape memories, or any other medium that can be used to carry or store data and is computer-readable.

[0104] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent in such a process, method, commodity or device. Without further limitations, the element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the element.

Claims

1. A motherboard testing method, characterized in that, The described test method includes the following steps: S1: Apply a pre-calibrated sweep signal to the motherboard to be tested, obtain impedance data of each test probe at different frequency points, and generate an impedance compensation matrix; S2: Apply a test excitation signal to the motherboard to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency domain conversion on all the original signals of the test probes, and then perform frequency domain calibration using the impedance compensation matrix to obtain the frequency domain signals of each calibrated test probe; S3: Perform multi-channel frequency domain filtering and linear combination processing on the frequency domain signals of each calibrated test probe to obtain the frequency domain signals of each test parameter and extract frequency domain features, convert the frequency domain signals of each test parameter into time domain signals and extract time domain features, and calculate the test parameter weight vector based on the time domain features and frequency domain features; S4: According to the preset probe weight mapping matrix, convert the test parameter weight vector into a test probe weight vector; S5: Based on the test probe weight vector, adjust the excitation intensity and sampling rate of each test probe through a preset adjustment rule.

2. The motherboard testing method according to claim 1, characterized in that, In step S1, the frequency range of the pre-calibrated sweep signal covers 1.5 times the fundamental frequency of the motherboard.

3. A motherboard testing method according to claim 2, characterized in that, The calculation formula for the impedance compensation matrix is: Among them, K is the impedance compensation matrix, and k n,f is the impedance compensation coefficient of the test probe n at the frequency point f, is a set of complex numbers, N is the total number of test probes, F is the total number of frequency points, and Z ref (f) is the preset reference impedance of the test probe at the frequency point f, |Z ref (f)| is the amplitude of the preset reference impedance of the test probe at the frequency point f, Z n (f) is the actual impedance of the test probe n at the frequency point f, |Z n (f)| is the amplitude of the actual impedance of the test probe n at the frequency point f, ∠Z n (f) is the phase of the actual impedance of the test probe n at the frequency point f, ∠Z ref (f) is the phase of the preset reference impedance of the test probe at the frequency point f, V n (f) is the complex voltage of the test probe n at the frequency point f, I n (f) is the complex current of the test probe n at the frequency point f, and j is the imaginary unit.

4. A motherboard testing method according to claim 3, wherein In step S2, the formula for performing frequency domain conversion on all the original signals of the test probes and then performing frequency domain calibration using the impedance compensation matrix is: Among them, S n (f) is the original signal of test probe n at frequency point f, is the calibrated signal of test probe n at frequency point f.

5. A motherboard testing method according to claim 4, characterized in that, In step S3, the generation steps of the test parameter weight vector are: Perform multi-channel frequency domain filtering and linear combination processing on the frequency domain signals of each calibrated test probe to obtain the frequency domain signals of each test parameter; Extract the frequency domain energy of the key frequency points of the frequency domain signals of each test parameter within the preset sensitive frequency band as the frequency domain feature; Perform inverse Fourier transform on the frequency domain signals of each test parameter, obtain the time domain signals of each test parameter, and calculate the time domain variance as the time domain feature; Based on the frequency domain features and time domain features, calculate the importance coefficients of each test parameter, and normalize the obtained importance coefficients of all test parameters to obtain the test parameter weight vector.

6. A motherboard testing method according to claim 5, characterized in that, The calculation formula for the importance coefficient is: Among them, R p is the importance coefficient of the test parameter p, is the time-domain variance of the test parameter p, and E p (f s ) is the frequency-domain energy of the test parameter at the key frequency point f s , and λ is the time-domain weight coefficient.

7. A motherboard testing method according to claim 6, characterized in that The calculation formula for the test parameter weight vector is: μ = [μ1, μ2, ……, μ P ​ where, μ p is the test parameter weight of test parameter p, and μ is the test parameter weight vector.

8. A motherboard testing method according to claim 7, characterized in that The calculation formula for the probe weight mapping matrix is: M = [m p,n P×N ​ w = [ω1, ω2, ……, ω N ​ Among them, M is the probe weight mapping matrix, ω n is the test probe weight of the test probe n, P is the total number of test parameters, m p,n is the weight of the dependence degree of the test parameter p on the test probe n, μ p is the test parameter weight of the test parameter p, and w is the test probe weight vector.

9. A motherboard testing method according to claim 8, characterized in that, The calculation formula for the dependence weight of test parameter p on test probe n is: where m p,m is the dependence degree weight of the test parameter p on the test probe n, and D p (f) is the sensitivity of the test parameter p at the frequency point f.

10. A main board testing system is applied to a main board testing method as described in claims 1 to 9, and is characterized in that, The described test system includes: A probe impedance calibration module, which is used to apply a pre-calibrated sweep signal to the motherboard to be tested, obtain impedance data of each test probe at different frequency points, and generate an impedance compensation matrix; A frequency domain calibration module, which is used to apply a test excitation signal to the motherboard to be tested, obtain the original signals of the test probes within a preset sliding time window, perform frequency domain conversion on all the original signals of the test probes, and then perform frequency domain calibration using the impedance compensation matrix to obtain the frequency domain signals of each calibrated test probe; A multi-domain feature fusion module, which is used to perform multi-channel frequency domain filtering and linear combination processing on the frequency domain signals of each calibrated test probe to obtain the frequency domain signals of each test parameter and extract frequency domain features, convert the frequency domain signals of each test parameter into time domain signals and extract time domain features, and calculate the test parameter weight vector based on the time domain features and frequency domain features; A weight mapping conversion module, configured to convert a test parameter weight vector into a test probe weight vector according to a preset probe weight mapping matrix; An excitation regulation module, configured to adjust the excitation intensity and sampling rate of each test probe based on the test probe weight vector through a preset adjustment rule.

Images