An online performance testing method for PCBA circuit boards and PCBA circuit boards

Abstract

This invention relates to the field of performance testing technology, and discloses an online performance testing method for PCBA circuit boards and a PCBA circuit board. The method involves: synchronously applying excitation current to an array of probes to establish an eddy current induction field covering the surface layer, solder joint layer, and inner layers of the copper conductors in the PCBA board under test, obtaining induced voltage signals at each probe location; digitally demodulating the induced voltage signals to obtain a set of impedance measurements at each probe location; constructing a first impedance vector based on the impedance measurement set and subtracting it from the reference impedance vector of a good product to obtain a second impedance vector; calculating the target phase angle, determining the defect type based on the target phase angle, and performing online rejection decisions. This invention solves the technical problem that existing single-frequency eddy current detection technology cannot achieve synchronous induction of multi-level defects due to fixed penetration depth, and realizes single-scan full-layer coverage detection of surface copper conductor breaks, cold solder joints, and via wall cracks.

Description

Technical Field

[0001] This invention relates to the field of performance testing technology, and in particular to an online performance testing method for PCBA circuit boards and a PCBA circuit board. Background Technology

[0002] Currently, PCBA production lines use AOI (Automated Optical Inspection) technology for online quality inspection of circuit boards. Existing technology uses CCD cameras to optically image and compare the board surface morphology, but can only identify visible surface defects such as missing solder joints and solder bridges. It cannot effectively detect hidden defects inside the copper conductors. Furthermore, PCBA circuit boards are highly susceptible to internal hidden defects during production, such as cold solder joints, via wall cracks, and broken surface copper conductors. These defects are completely identical to good products in terms of board morphology, and AOI technology may miss them, leading to defective products flowing downstream and causing batch failures. Summary of the Invention

[0003] The main objective of this invention is to provide an online performance testing method for PCBA circuit boards and a PCBA circuit board. This invention solves the technical problem that existing single-frequency eddy current testing technology cannot achieve synchronous sensing of multi-level defects due to the fixed penetration depth, and realizes full-level coverage detection of surface copper wire breaks, cold solder joints, and via wall cracks in a single scan.

[0004] To achieve the above objectives, the present invention provides an online performance testing method for PCBA circuit boards, comprising the following steps: Excitation current is synchronously applied to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer and inner layer in the copper conductor of the PCBA board under test, and obtain the induced voltage signal at each probe position. The induced voltage signal is digitally phase-sensitive demodulated to obtain a set of impedance measurements at each probe position. A first impedance vector is constructed based on the impedance measurement set, and a second impedance vector is obtained by subtracting the first impedance vector from the reference impedance vector of the good product. The target phase angle is calculated based on the second impedance vector, the defect type is determined based on the target phase angle, and an online rejection decision is executed.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of synchronously applying excitation current to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer, and inner layer in the copper conductor of the PCBA board under test, and obtaining the induced voltage signal at each probe position, includes: The array probes are scanned row by row and column coordinates, and an excitation current superimposed by three excitation frequencies is applied simultaneously. Based on the eddy current penetration depth corresponding to each excitation frequency, an eddy current induction field covering the surface layer, solder joint layer and inner layer is established in the copper conductor of the PCBA board under test. Based on the eddy current sensing field, the induced voltage signal at each probe position is collected.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of acquiring the induced voltage signal at each probe position based on the eddy current induction field includes: Based on the electromagnetic induction between the eddy current induction field and the copper conductor of the PCBA board under test, an induction port voltage carrying information on the impedance change of the copper conductor is generated at the induction port of each probe coil. The voltage at the sensing port is input to an analog-to-digital converter for digital processing to obtain the induced voltage signal at each probe position.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the step of digitally demodulating the induced voltage signal to obtain a set of impedance measurements at each probe position includes: The induced voltage signal is multiplied by the in-phase reference signal and the quadrature reference signal corresponding to each excitation frequency to obtain the in-phase filter component and the quadrature filter component of each excitation frequency. Within an integer multiple of the complete cycle of each excitation frequency, a discrete integration operation is performed on the in-phase filter component and the quadrature filter component to extract the impedance measurement set for each probe position.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the step of performing phase-sensitive multiplication operations on the induced voltage signal with the in-phase reference signal and the quadrature reference signal corresponding to each excitation frequency to obtain the in-phase filter component and the quadrature filter component of each excitation frequency includes: Based on the excitation frequencies output by the direct digital frequency synthesis module, in-phase reference signals and quadrature reference signals corresponding to the excitation current are generated in the digital domain. The induced voltage signal is multiplied point by point with the in-phase reference signal and the quadrature reference signal to obtain the in-phase multiplication integral and the quadrature multiplication integral corresponding to each excitation frequency; The in-phase multiplicative integral and the quadrature multiplicative integral are respectively filtered by a low-pass filter to remove the second harmonic component and the frequency aliasing term, so as to obtain the in-phase filtered component and the quadrature filtered component of each excitation frequency.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of performing discrete integration on the in-phase filter component and the quadrature filter component within an integer multiple of the complete cycle of each excitation frequency to extract the impedance measurement set at each probe position includes: Within the integration interval of an integer multiple of the excitation frequency, a discrete integration operation is performed on the in-phase filter component corresponding to each excitation frequency to obtain the real component of the first impedance corresponding to each excitation frequency. Within the integration interval of an integer multiple of the excitation frequency, a discrete integration operation is performed on the orthogonal filter components corresponding to each excitation frequency to obtain the first imaginary impedance component corresponding to each excitation frequency. The real component of the first impedance corresponding to each excitation frequency is combined with the imaginary component of the first impedance to obtain the impedance measurement set for each probe position.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the step of constructing a first impedance vector based on the impedance measurement set and subtracting the first impedance vector from the reference impedance vector of a good product to obtain a second impedance vector includes: Based on the preset lift-off sensitivity vector, the first real component and the first imaginary component of the impedance corresponding to each excitation frequency in the impedance measurement group are lifted off and compensated respectively to obtain the second real component and the second imaginary component of the impedance. The real component of the second impedance is combined with the imaginary component of the second impedance to form a first impedance vector; The first impedance vector is subtracted from the reference impedance vector of the good product to obtain the real part difference component and the imaginary part difference component of the impedance corresponding to each excitation frequency. The real part difference component and the imaginary part difference component of the impedance are then combined to form the second impedance vector.

[0011] Optionally, in a seventh implementation of the first aspect of the present invention, the step of calculating the target phase angle based on the second impedance vector, determining the defect type based on the target phase angle, and performing an online rejection decision includes: On the complex impedance plane corresponding to each excitation frequency, the four-quadrant arctangent operation is performed on the difference components of the real part of the impedance and the difference components of the imaginary part of the impedance corresponding to each excitation frequency in the second impedance vector to obtain the target phase angle corresponding to each excitation frequency. The defect type is determined based on the target phase angle, and the comprehensive defect severity index is calculated using the reference impedance mode corresponding to each excitation frequency. An online rejection decision is then made for the PCBA board under test based on the comprehensive defect severity index.

[0012] Optionally, in an eighth implementation of the first aspect of the present invention, the step of determining the defect type based on the target phase angle, calculating the comprehensive defect severity index using the reference impedance modulus corresponding to each excitation frequency, and performing an online rejection decision on the PCBA board under test based on the comprehensive defect severity index includes: The defect type is determined based on whether the amplitude of the target phase angle corresponding to each excitation frequency and the amplitude of the corresponding impedance real part difference component exceed the corresponding amplitude threshold. Based on the real part difference component and the imaginary part difference component of the impedance, the first difference component magnitude and the second difference component magnitude of the real part difference component of the impedance corresponding to each excitation frequency in the second impedance vector are calculated respectively. A weighted summation is performed on the first difference component magnitude and the second difference component magnitude to obtain the comprehensive defect severity index; The comprehensive defect severity index is compared with the severity threshold; If the overall defect severity index exceeds the severity threshold, an online rejection decision is triggered for the PCBA board under test.

[0013] The present invention also provides a PCBA circuit board for implementing the steps of the method described in any of the above embodiments.

[0014] In summary, this invention solves the technical problem of existing single-frequency eddy current detection technology, which cannot achieve synchronous sensing of multi-level defects due to fixed penetration depth, by simultaneously applying excitation currents superimposed by three excitation frequencies to the array probe. It establishes an eddy current induction field covering the surface layer, solder joint layer, and inner layer of the copper conductor in the PCBA board under test based on the eddy current penetration depth corresponding to each excitation frequency. This enables single-scan full-layer coverage detection of surface copper conductor breaks, cold solder joints, and via wall cracks. By performing three-channel parallel digital phase-sensitive demodulation on the induced voltage signal, and using the excitation frequency of the direct digital frequency synthesis module as a reference, phase-aligned in-phase and quadrature reference signals are generated in the digital domain. After point-by-point multiplication, low-pass filtering, and discrete integration within integer multiples of the complete cycle, the real and imaginary components of the first impedance corresponding to each excitation frequency are accurately extracted, eliminating impedance measurement errors caused by inter-channel crosstalk and temperature drift in traditional analog demodulation schemes. By performing lift-off compensation processing based on the lift-off sensitivity vector on the impedance measurement group and subtracting it from the good product reference impedance vector, a second impedance vector that only reflects the impedance deviation caused by defects is obtained, eliminating the masking of defect signals by the background impedance of high-density traces. Furthermore, based on the target phase angle interval of the second impedance vector on the complex impedance plane, defect judgment rules are matched step by step according to the priority order of via cracks, surface wire breaks, and cold solder joints. The comprehensive defect severity index is calculated with the reference impedance modulus as the normalized weight to execute the rejection decision. This realizes the online full-loop inspection of surface copper wire breaks, cold solder joints, and via wall cracks within the PCBA production line cycle. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the steps of an online performance testing method for PCBA circuit boards in one embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the steps for obtaining the induced voltage signal at each probe position in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the steps of digital phase-sensitive demodulation in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the steps of constructing an impedance vector in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the steps for determining the defect type in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the online elimination decision-making steps in an embodiment of the present invention.

[0016] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Reference Figure 1 This embodiment provides an online performance testing method for PCBA circuit boards, including the following steps: S1, apply excitation current synchronously to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer and inner layer in the copper conductor of the PCBA board under test, and obtain the induced voltage signal at each probe position; S2, digital phase-sensitive demodulation of the induced voltage signal to obtain the impedance measurement value set for each probe position; S3, construct a first impedance vector based on the impedance measurement value group, and subtract the first impedance vector from the reference impedance vector of the good product to obtain the second impedance vector; S4. Calculate the target phase angle based on the second impedance vector, determine the defect type based on the target phase angle, and execute an online rejection decision.

[0019] In this embodiment, by synchronously applying excitation currents superimposed by three excitation frequencies to the array probes, an eddy current induction field covering the surface layer, solder joint layer, and inner layer of the copper conductor in the PCBA board under test is simultaneously established based on the eddy current penetration depth corresponding to each excitation frequency. This solves the technical problem that existing single-frequency eddy current detection technology cannot achieve synchronous induction of multi-level defects due to the fixed penetration depth, and realizes full-level coverage detection of surface copper wire breaks, cold solder joints, and via wall cracks in a single scan. By performing three-channel parallel digital phase-sensitive demodulation on the induced voltage signal, and using the excitation frequency of the direct digital frequency synthesis module as a reference, phase-aligned in-phase reference signals and quadrature reference signals are generated in the digital domain. After point-by-point multiplication, low-pass filtering, and discrete integration within integer multiples of the complete cycle, the real and imaginary components of the first impedance corresponding to each excitation frequency are accurately extracted, eliminating impedance measurement errors caused by inter-channel crosstalk and temperature drift in traditional analog demodulation schemes. By performing lift-off compensation processing based on the lift-off sensitivity vector on the impedance measurement group and subtracting it from the good product reference impedance vector, a second impedance vector that only reflects the impedance deviation caused by defects is obtained, eliminating the masking of defect signals by the background impedance of high-density traces. Furthermore, based on the target phase angle interval of the second impedance vector on the complex impedance plane, defect judgment rules are matched step by step according to the priority order of via cracks, surface wire breaks, and cold solder joints. The comprehensive defect severity index is calculated with the reference impedance modulus as the normalized weight to execute the rejection decision. This realizes the online full-loop inspection of surface copper wire breaks, cold solder joints, and via wall cracks within the PCBA production line cycle.

[0020] In one example, such as Figure 2 Excitation current is synchronously applied to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer, and inner layer in the copper conductor of the PCBA board under test, and to obtain the induced voltage signal at each probe position, including: S11, scan the array probe row by row according to the column coordinates, and simultaneously apply an excitation current superimposed by three excitation frequencies. Based on the eddy current penetration depth corresponding to each excitation frequency, establish an eddy current induction field covering the surface layer, solder joint layer and inner layer in the copper conductor of the PCBA board under test. S12, based on the eddy current induction field, collects the induced voltage signal at each probe position.

[0021] In this embodiment, after the PCBA board to be tested enters the online inspection station, an array probe composed of N rows × M columns of planar micro coils is installed below the gantry scanning mechanism, and the row and column coordinate systems of the array probes are mechanically aligned with the coordinate system of the PCBA board surface. The center-to-center distance between adjacent coils of the array probe and the row-by-row step distance can be set to 0.5 mm. This value is suitable for scenarios where the preset minimum line width, pad spacing, and via spacing of the target board all meet the inspection resolution requirements. For higher density trace areas, the control unit can reduce the step distance according to the board layout data, or use probe coils with smaller effective sensing diameters to ensure that the scanning trajectory covers the area where the target defect is located. The probe lift-off height is set to... This height balances the safety gap between the probe coil and the board surface with the eddy current coupling strength, preventing the probe from contacting the solder mask layer while keeping the amplitude of the induced signal within the demodulation range. The scanning control unit performs row-by-row step-by-step scanning of the array probes according to the row and column coordinates, with a single step distance set to 0.5mm to ensure no obvious detection gaps between adjacent scan rows, and synchronously triggers the excitation drive circuit at each probe coordinate position. The excitation drive circuit applies an excitation current to each probe coil, which is formed by the superposition of three frequency components, and the excitation current satisfies the formula: ; in, This represents the total excitation current applied to the probe coil at time t, in A. , , These represent the current amplitudes of the first, second, and third frequency components, respectively, and can be calibrated based on the number of turns of the probe coil, the lift-off height, the distribution of sensitive devices on the board, and the allowable induced voltage range. For example, the current amplitude can be set to 50mA and used when the PCBA board under test is not powered on, the probe is in non-contact coupling with the board surface, and the onboard magnetic sensitive devices and high-impedance analog input areas have undergone safety verification. , , These represent the first, second, and third excitation frequencies, for example, 500kHz, 2MHz, and 5MHz respectively. This represents the sampling or excitation time, measured in seconds (s). The three frequency components are generated in the digital domain by a direct digital frequency synthesis module, and their initial phases are synchronously cleared. After digital-to-analog conversion and power drive, they are input to the array probe coils, ensuring that each probe position receives phase-consistent excitation. The eddy current penetration depth corresponding to each frequency is calculated using the formula: ; in, This represents the eddy current penetration depth corresponding to the nth excitation frequency, in meters (m). This indicates the frequency index, which can be 1, 2, or 3. To represent the permeability of a copper conductor, take... ; To represent the conductivity of a copper conductor, take... Based on the above parameters, the penetration depth corresponding to 500kHz is approximately 92μm, the penetration depth corresponding to 2MHz is approximately 46μm, and the penetration depth corresponding to 5MHz is approximately 29μm. Thus, a layered eddy current induction field is simultaneously formed in the copper conductor of the PCBA board under test, covering the inner layer via wall, the main area of ​​the solder joint, and the surface copper conductor.

[0022] Based on this layered eddy current induction field, electromagnetic coupling is generated between the probe coil and the copper conductor of the PCBA board under test. The impedance changes caused by the continuity of the copper conductor, the internal bonding state of the solder joint, and the integrity of the via wall are converted into voltage responses at the induction ports of each probe coil. The voltage at the induction port at each row and column coordinate position is synchronously acquired according to the scan trigger sequence, and the induction voltage signal is associated and stored with the probe coordinate, scan row number, and acquisition timestamp to form the original induction voltage signal sequence.

[0023] In one example, S12 includes: Based on the electromagnetic induction between the eddy current induction field and the copper conductor of the PCBA board under test, an induction port voltage carrying the impedance change information of the copper conductor is generated at the induction port of each probe coil. The voltage at the sensing port is input to an analog-to-digital converter for digital processing to obtain the induced voltage signal at each probe position.

[0024] In this embodiment, after the array probe completes the positioning of the current row and column coordinates, the excitation drive circuit maintains the phase continuity of the three-frequency superimposed excitation, so that a stable alternating magnetic field is formed around the probe coil. After the alternating magnetic field enters the copper conductor of the PCBA board under test, it induces eddy current distribution in the surface copper wire, the metal connection area of ​​the solder joint, and the copper layer of the via wall. When the copper conductor is continuous, the eddy current path is relatively complete, and the probe coil sensing port exhibits a complex impedance response close to that of a good product reference. When the surface copper wire is broken, the cold solder interface of the solder joint is discontinuous, or cracks appear in the via wall, the local current closed path and magnetic field reaction change, the equivalent impedance of the probe coil changes accordingly, and a weak AC voltage carrying impedance change information is formed at the sensing port. To ensure the stable input of the sensing port voltage into the digital phase-sensitive demodulation link, a differential input conditioning unit is installed after the sensing port of each probe coil. The common-mode rejection ratio (CMRR) of the differential amplifier is set to no less than 80dB to suppress common-mode interference generated by the production line motor, switching power supply, and synchronous excitation of adjacent coils. The amplification gain is set to 40dB, corresponding to a voltage gain of approximately 100 times, which raises the millivolt-level sensing port voltage to the effective input range of the analog-to-digital converter (ADC). Simultaneously, a limiting protection and input impedance matching network are configured at the front end of the amplification channel to reduce overshoot caused by probe switching, local board warping, or transient coupling abrupt changes. The conditioned sensing port voltage is then synchronously sampled by a 16-bit ADC at a sampling rate of 50MSPS. This sampling rate forms a 10-fold oversampling relationship relative to the highest excitation frequency of 5MHz, which helps retain phase information and reduce discrete sampling errors. The ADC sampling clock and the excitation signal generation clock use the same reference crystal oscillator, ensuring a fixed correspondence between the sampling phase and the excitation phase and avoiding random phase drift caused by asynchronous sampling. During the digitization process, the control unit uses the probe row and column coordinates, the scan trigger signal, and the sampling window number as indexes to write the voltage sample sequence collected at each probe position within the dwell time into the scan data buffer. It also marks the status of saturated samples that significantly exceed the input range and generates induced voltage signals for unsaturated sample sequences that are clock-synchronized.

[0025] In one example, such as Figure 3 The induced voltage signal is digitally phase-sensitively demodulated to obtain a set of impedance measurements at each probe position, including: S21, perform phase-sensitive multiplication operations on the induced voltage signal with the in-phase reference signal and the quadrature reference signal corresponding to each excitation frequency to obtain the in-phase filter component and the quadrature filter component of each excitation frequency. S22, within the complete cycle of integer multiples of each excitation frequency, performs discrete integration on the in-phase filter component and the quadrature filter component to extract the impedance measurement set at each probe position.

[0026] In this embodiment, after the induced voltage signal at each probe position completes analog-to-digital conversion, the scanning control unit sends the digital voltage sample sequence corresponding to the same probe position into three parallel digital phase-sensitive demodulation channels according to the probe row and column coordinates, sampling window number, and scanning trigger time. The three demodulation channels correspond to the first excitation frequency of 500kHz, the second excitation frequency of 2MHz, and the third excitation frequency of 5MHz, respectively. The first excitation frequency corresponds to a deeper equivalent sensing region and is mainly used to reflect the impedance disturbances of the copper layer near the via, the deeper pad connection area, and the larger volume metal connection structure. The second excitation frequency corresponds to the middle equivalent sensing region and is mainly used to reflect the impedance disturbances of the main solder joint area and the area near the solder interface. The third excitation frequency corresponds to a shallower equivalent sensing region and is mainly used to reflect the impedance disturbances of the surface copper conductor and the near-surface copper foil. To avoid the three frequency components from mixing in the superimposed induced voltage signal, the direct digital frequency synthesis module generates the three-frequency excitation signal and simultaneously provides the demodulation unit with in-phase and quadrature reference signals that are consistent with the phase of each excitation frequency. The demodulation unit sends the induced voltage sample sequence into the in-phase channel and the quadrature channel respectively, and performs point-by-point phase-sensitive multiplication processing according to the same sampling clock, so that the target frequency component is converted into a low-frequency correlated component, while other non-target frequency components, second harmonic components and incoherent interference in the production line environment remain as filterable AC disturbances. Each demodulation channel first performs coherent accumulation processing on the product result of the in-phase channel and the quadrature channel within the complete cycle window of the corresponding excitation frequency. This process preserves the relevant components corresponding to the target excitation frequency and weakens non-target frequency components, residual second harmonic components, and incoherent interference during accumulation. After the coherent accumulation processing is completed, the accumulation result is input into a low-pass smoothing link for digital filtering to obtain in-phase and quadrature filter components that are coherent with the target excitation frequency. This reduces common problems in analog phase-locked loops such as inter-channel crosstalk, temperature drift, and phase inconsistency.

[0027] After low-pass smoothing, the control unit performs discrete integration within integer multiples of the full cycle window corresponding to each excitation frequency. The first excitation frequency is for 10 full cycles, the second for 40, and the third for 100. The effective integration time for all three channels is aligned to 20 μs, ensuring the three impedance responses correspond to the same probe dwell position and the same sampling window. When filtering transients or group delays exist in the low-pass smoothing link, the control unit determines the effective output sample range based on the filter order and removes transient samples during discrete integration to avoid impedance component deviations caused by filter instability. After discrete integration, the in-phase channel outputs the real impedance component characterizing resistive changes, while the positive-phase channel outputs the imaginary impedance component characterizing inductive reactance and eddy current coupling changes. The control unit then combines the real and imaginary impedance components corresponding to the three excitation frequencies at the same probe position in a fixed order to form a three-frequency, six-dimensional impedance measurement set for each probe position.

[0028] In one example, S21 includes: Based on the excitation frequencies output by the direct digital frequency synthesis module, in-phase reference signals and quadrature reference signals corresponding to the excitation current are generated in the digital domain. The induced voltage signal is multiplied point by point with the in-phase reference signal and the quadrature reference signal to obtain the in-phase multiplication integral and the quadrature multiplication integral corresponding to each excitation frequency; The in-phase and quadrature multiplicative integrals are filtered by a low-pass filter to remove the second harmonic component and frequency aliasing term, respectively, to obtain the in-phase and quadrature filtered components for each excitation frequency.

[0029] In this embodiment, the digital phase-sensitive demodulation unit uses the three excitation frequencies output by the direct digital frequency synthesis module as phase references, and establishes three parallel demodulation channels for 500kHz, 2MHz, and 5MHz respectively. The three demodulation channels share the same reference clock with the excitation current, maintaining a fixed correspondence between the sampling phase and the excitation phase. For the first... A single excitation frequency is used to generate in-phase and quadrature reference signals in the digital domain. The reference signals satisfy the following relationship: ; ; in, Indicates the first The excitation frequency at the ... In-phase reference values ​​at each sampling point Indicates the first The excitation frequency at the ... Orthogonal reference values ​​at each sampling point Indicates the first One excitation frequency, Choosing 1, 2, and 3 corresponds to 500kHz, 2MHz, and 5MHz respectively. Indicates the sampling point number. This represents the sampling rate of the analog-to-digital converter, set to 50 MSPS. The in-phase and quadrature reference signals mentioned above are synchronously reset to their initial phase values ​​by the direct digital frequency synthesis module at each scan trigger, ensuring that the demodulation starting point remains consistent across different probe coordinates.

[0030] The induced voltage digital sequence acquired at the same probe coordinates is sent to the in-phase channel and the quadrature channel respectively, and multiplied point by point with the reference signal of the corresponding frequency to obtain the in-phase multiplication integral and the quadrature multiplication integral corresponding to each excitation frequency. The calculation relationship is as follows: ,in, Indicates the first The in-phase multiplication integral corresponding to each excitation frequency Indicates the first The orthogonal integral quantity corresponding to each excitation frequency Indicates the probe coil at the 1st Digital samples of induced voltage obtained from each sampling point. Through point-by-point phase-sensitive multiplication, the target component in the induced voltage signal that is related to the frequency and phase of the reference signal is converted into a low-frequency correlated component, while other frequency components, second harmonic components, and incoherent interference still appear as AC disturbances.

[0031] The in-phase and quadrature multiplication integrals are respectively input to a finite impulse response low-pass filter with a cutoff frequency of 10kHz. The low-pass filter suppresses the second harmonic component, frequency aliasing terms, and residual components from adjacent excitation frequencies, while retaining stable components coherent with the target excitation frequency. Each demodulation channel outputs the in-phase and quadrature filtered components corresponding to the excitation frequency.

[0032] In one example, S22 includes: Within the integration interval of an integer multiple of the excitation frequency and the complete period, a discrete integration operation is performed on the in-phase filter component corresponding to each excitation frequency to obtain the real component of the first impedance corresponding to each excitation frequency. Within the integration interval of an integer multiple of the excitation frequency and the complete period, a discrete integration operation is performed on the orthogonal filter components corresponding to each excitation frequency to obtain the first imaginary component of the impedance corresponding to each excitation frequency. By combining the real component and the imaginary component of the first impedance corresponding to each excitation frequency, the impedance measurement set for each probe position is obtained.

[0033] In this embodiment, the integration interval is defined as an integer multiple of the excitation frequency, ensuring that the start and end points of the integration window fall within the same phase state. This reduces the impact of incoherent noise, residual second harmonic components, and sampling window truncation errors on the impedance results. The integration interval corresponding to the first excitation frequency of 500kHz is 10 complete cycles, the integration interval corresponding to the second excitation frequency of 2MHz is 40 complete cycles, and the integration interval corresponding to the third excitation frequency of 5MHz is 100 complete cycles. The integration time for all three is 20μs. This time setting ensures that the impedance extraction of the three frequency channels is completed within the same probe dwell window, avoiding spatial mismatch caused by position changes during the array probe's row-by-row scanning process. After performing discrete integration on the in-phase filtered components, the real component of the first impedance corresponding to each excitation frequency is obtained; after performing discrete integration on the quadrature filtered components, the imaginary component of the first impedance corresponding to each excitation frequency is obtained. The calculation relationship can be expressed as: ; ; in, Indicates the first The real component of the first impedance corresponding to each excitation frequency mainly reflects the resistive response caused by the continuous change of the copper conductor. Indicates the first The imaginary component of the first impedance corresponding to each excitation frequency mainly reflects the eddy current coupling strength and inductive reactive response; Indicates the first The integral interval corresponding to each excitation frequency; This represents the in-phase component after low-pass filtering; This represents the orthogonal components after low-pass filtering. After the discrete integration operation is completed, the control unit sequentially writes the real and imaginary components of the first impedance obtained at the first, second, and third excitation frequencies into the same probe coordinate data structure in ascending order of frequency, forming a set of impedance measurement values ​​for each probe position. , , , , , The impedance characteristics consist of three frequencies and six dimensions. This set of impedance measurements is then synchronously bound to the probe row and column coordinates, scan row number, and sampling window number and stored in the current scan data buffer.

[0034] In one example, such as Figure 4 A first impedance vector is constructed based on the impedance measurement values. The second impedance vector is obtained by subtracting the first impedance vector from the reference impedance vector of a good product. This second impedance vector includes: S31, based on the preset lift-off sensitivity vector, lift-off compensation is performed on the first real part and the first imaginary part of the impedance corresponding to each excitation frequency in the impedance measurement group to obtain the second real part and the second imaginary part of the impedance. S32, combine the real part of the second impedance with the imaginary part of the second impedance to form the first impedance vector; S33, the first impedance vector is subtracted from the reference impedance vector of the good product to obtain the real part difference component and the imaginary part difference component of the impedance corresponding to each excitation frequency, and the real part difference component and the imaginary part difference component of the impedance are combined into the second impedance vector.

[0035] In this embodiment, the control unit calls the lift-off sensitivity data corresponding to the current PCBA board model, batch number, and probe row and column coordinates, and inputs the first real impedance component and the first imaginary impedance component from the impedance measurement value group into the lift-off compensation link. Because the array probes are affected by slight board warping, conveyor belt flatness errors, and minor vibrations of the probe frame during online scanning, the actual lift-off height between the probe coil and the PCBA board surface may deviate from the standard lift-off height. This deviation will simultaneously change both the real and imaginary impedance components. If the impedance measurement value is directly compared with the good product reference value, the lift-off disturbance may be misjudged as a broken copper wire, cold solder joint, or via wall crack.

[0036] The lift-off deviation at the current coordinate position is estimated based on a preset lift-off sensitivity vector. For example, the preset lift-off sensitivity vector can be obtained during the baseline establishment phase through two scans at a standard lift-off height of 0.3 mm and an offset height of 0.35 mm. The offset is set to 0.05 mm, a value that facilitates a sufficiently significant impedance change within the controllable range of the mechanical pad, while not disrupting the safe clearance between the probe and the board surface. For the nth excitation frequency, the lift-off sensitivity vector is expressed as: ; in, Representing coordinates First The lift-off sensitivity vector corresponding to each excitation frequency, in units of... ; This represents the change in the real part of the impedance between the offset height scan and the standard height scan, in units of... ; This represents the change in the imaginary part of impedance between the offset height scan and the standard height scan, in units of... ; This represents the liftoff height offset applied during the baseline establishment phase, taken as 0.05 mm; and These represent the row and column coordinates corresponding to the probe array, respectively. Pick , , These correspond to 500kHz, 2MHz, and 5MHz, respectively. During online testing, the control unit performs a preliminary difference comparison between the current impedance measurement set and the good product reference impedance value, and projects the difference vector along the direction of the lift-off sensitivity vector to estimate the actual lift-off deviation of the current probe coordinate position. Subsequently, the impedance change component introduced by the lift-off deviation is subtracted from the first impedance real component and the first impedance imaginary component to obtain the second impedance real component and the second impedance imaginary component. The corresponding compensation relationship can be expressed as: ; ; in, Indicates the first before compensation The real component of the first impedance at each excitation frequency, in units of ; Indicates the first before compensation The imaginary component of the first impedance at each excitation frequency, in units of ; Indicates the compensation after the first The real component of the second impedance at each excitation frequency, in units of ; Indicates the compensation after the first The imaginary component of the second impedance at each excitation frequency, in units of ; This represents the current liftoff deviation estimated during online detection, in mm. This represents the component in the lift-off sensitivity vector corresponding to the real part of the impedance, in units of... ; This represents the component in the lift-off sensitivity vector corresponding to the imaginary part of the impedance, in units of... .

[0037] After lift-off compensation is completed, the real and imaginary components of the second impedance are combined into a first impedance vector according to the first, second, and third excitation frequencies. This ensures that the compensated impedance responses at the same probe position across the three depth levels form a unified data structure. The first impedance vector can be expressed as: ; in, Representing coordinates The first impedance vector at point 1 has six components, all of which are impedance components after lift-off compensation, with units of mΩ.

[0038] The reference impedance vectors at the same coordinate position are read from the good product reference impedance database, and the first impedance vector is subtracted from the reference impedance vector of the good product component by component to obtain the difference characteristics that only reflect the impedance deviation caused by defects. The difference calculation relationship can be expressed as: ; in, Representing coordinates The second impedance vector at that location; Representing coordinates The good reference impedance vector at the location; Indicates the first The real part difference component of the impedance corresponding to each excitation frequency, in mΩ; Indicates the first The impedance difference component corresponding to each excitation frequency is expressed in mΩ. The background impedance formed by the board trace density, copper foil width, pad geometry, and normal stack-up structure is canceled out by the good product reference impedance vector. The non-defect disturbances caused by the change in probe lift-off height are also weakened by the lift-off compensation process. The resulting second impedance vector can more effectively characterize the difference between the real and imaginary impedance caused by broken surface copper conductors, cold solder joints, and via wall cracks.

[0039] In one example, such as Figure 5 The target phase angle is calculated based on the second impedance vector. Based on the target phase angle, the defect type is determined and an online rejection decision is executed, including: S41, On the complex impedance plane corresponding to each excitation frequency, perform four-quadrant arctangent operation on the difference components of the real part of the impedance and the difference components of the imaginary part of the impedance corresponding to each excitation frequency in the second impedance vector to obtain the target phase angle corresponding to each excitation frequency. S42 determines the defect type based on the target phase angle, calculates the comprehensive defect severity index using the reference impedance modulus corresponding to each excitation frequency, and makes an online rejection decision for the PCBA board under test based on the comprehensive defect severity index.

[0040] In this embodiment, the data processing unit uses the difference components of the real and imaginary parts of the impedance at the same probe coordinate position corresponding to three excitation frequencies as input, and establishes difference characteristic coordinate systems for the first, second, and third excitation frequencies respectively in the complex impedance plane. The horizontal coordinate corresponds to the difference component of the real part of the impedance, and the vertical coordinate corresponds to the difference component of the imaginary part of the impedance. Since the difference component of the real part of the impedance reflects the continuity change of the copper conductor, the change of resistive loss, and the degree of obstruction of the local conduction path, and the difference component of the imaginary part of the impedance reflects the eddy current coupling state, inductive reactance response, and changes in conductor geometric distribution, different defects will form difference vectors with different directions in the complex impedance plane. For surface copper conductor breaks, the shallow eddy current response corresponding to higher excitation frequencies is more pronounced, and the difference vector in the plane containing the third excitation frequency shows a stronger change in the real part and a relatively limited change in the imaginary part. For cold solder joints, lower excitation frequencies can cover the main area of ​​the solder joint, and the difference vector in the plane containing the first excitation frequency will show changes in both the real and imaginary parts. For via wall cracks, all three frequencies may be affected by the decrease in the continuity of the via copper layer, and the difference vectors in the three sets of complex impedance planes will show a relatively consistent characteristic of increasing real parts. In order to convert the above directional differences into identifiable angular characteristics, the target phase angle corresponding to each excitation frequency is calculated using four-quadrant arctangent calculation, and the calculation relationship is as follows: ,in, Indicates the first The target phase angle corresponding to each excitation frequency has a range of values. 180° to +180°; Indicates the first The impedance difference component corresponding to each excitation frequency; Indicates the first The impedance real part difference component corresponding to each excitation frequency; Pick , , These correspond to 500kHz, 2MHz, and 5MHz respectively.

[0041] The control unit sends the target phase angle and the amplitude of the difference between the real part of the impedance at the corresponding frequency to the defect judgment link. Based on the preset phase angle ranges entered by the three target phase angles, whether the corresponding real part difference components of the impedance exceed preset amplitude thresholds, and the combination relationship between the three frequency responses, the anomaly type at the current probe coordinate of the PCBA board under test is judged. Specifically, via wall cracks are primarily identified based on the synchronous increase of the real part difference between the three frequencies and the phase angles all deviating towards the real axis. Surface copper conductor breaks can be distinguished based on the shallow response of the third excitation frequency and the weaker response of the first excitation frequency. Cold solder joints can be identified based on the intermediate phase angle characteristics formed within the solder joint depth range of the first excitation frequency. After the defect type is identified, the reference impedance modulus corresponding to each excitation frequency in the good product reference impedance database is called to normalize the difference impedance amplitude at the current coordinate position, and a comprehensive defect severity index is formed accordingly. After the reference impedance modulus participates in the normalization, the background impedance differences caused by different trace widths, pad areas, via structures, and local copper densities on the board surface are weakened, making defects of the same severity have closer judgment scales in different coordinate regions. The control unit compares the comprehensive defect severity index with the preset severity threshold. When the current probe coordinate position meets both the defect type determination condition and the severity threshold condition, the PCBA board under test is marked as a defective product, and the defect coordinate, defect type, and severity result are written into the detection record and a rejection trigger signal is output to the online rejection mechanism. When the comprehensive defect severity index does not reach the threshold condition, the detection result is retained as monitoring data and no rejection action is triggered.

[0042] In one example, S42 includes: S421, Determine the defect type based on whether the amplitude of the target phase angle corresponding to each excitation frequency belongs to the preset phase angle interval and the amplitude of the corresponding impedance real part difference component exceeds the corresponding amplitude threshold. S422, based on the real part difference component and the imaginary part difference component of the impedance, calculate the first difference component magnitude of the real part difference component of the impedance and the second difference component magnitude of the imaginary part difference component of the impedance corresponding to each excitation frequency in the second impedance vector. S423, perform a weighted summation on the first difference component modulus and the second difference component modulus to obtain the comprehensive defect severity index; S424 compares the overall defect severity index with the severity threshold; S425 If the overall defect severity index exceeds the severity threshold, an online rejection decision for the PCBA board under test is triggered.

[0043] In this embodiment, the determination unit matches the target phase angle with a preset phase angle range and simultaneously reads the amplitude of the impedance real part difference component at the same frequency, forming a defect identification condition of "angle range + amplitude threshold + multi-frequency combination relationship". The impedance real part difference threshold and target phase angle range corresponding to each defect type are obtained by calibrating good samples, defect samples and repeated scan data of the same board type, and are stored according to board type number, copper thickness, pad size, via diameter and probe coordinate area respectively. For via wall cracks, the determination priority is set to the highest. When the impedance real part difference components corresponding to the first excitation frequency, the second excitation frequency and the third excitation frequency all exceed the via determination threshold obtained by the same board type, and the three target phase angles are all within the via determination range, it is determined to be a via wall crack. This numerical threshold is used to characterize that there is a significant resistive increase in the three frequency responses, which is suitable for distinguishing the discontinuity of the copper layer of the via that penetrates or is close to penetrating the depth direction. For surface copper conductor breaks, if the via wall crack condition is not met, and the target phase angle corresponding to the third excitation frequency is within the surface breakage judgment range, and the amplitude of the impedance real part difference component corresponding to the third excitation frequency exceeds the surface breakage threshold obtained from the same board type calibration, while the impedance real part difference component corresponding to the first excitation frequency is lower than the deep response exclusion threshold, then it is determined to be a surface copper conductor breakage. This condition utilizes the characteristic of strong high-frequency shallow response and weak low-frequency deep response to distinguish surface breakage from via wall cracks. For cold solder joints, if the via wall crack and surface copper conductor breakage judgment conditions are not met, the target phase angle corresponding to the second excitation frequency and the impedance real part difference component are used as the main criteria, combined with the response amplitude of the first excitation frequency for auxiliary confirmation. If the target phase angle corresponding to the second excitation frequency enters the cold solder joint preset range, and the impedance real part difference component corresponding to the second excitation frequency exceeds the cold solder joint amplitude threshold, then it is determined to be a cold solder joint. This range is used to characterize the synchronous changes of the real and imaginary parts caused by the discontinuity of the internal interface of the solder joint.

[0044] After the initial defect type determination, the processing unit calculates the differential impedance amplitude based on the difference between the real and imaginary impedance components at the same frequency, and normalizes it using the reference impedance modulus corresponding to each excitation frequency to obtain the comprehensive defect severity index. The calculation relationship is as follows: ; in, Representing coordinates The overall defect severity index at the location; , , These represent the difference components of the real part of the impedance corresponding to the three excitation frequencies; , , These represent the components of the impedance difference between the three excitation frequencies; , , These represent the reference impedance modes corresponding to the three excitation frequencies at the same coordinate position. After normalization using the reference impedance modes, the differences in background impedance caused by different trace widths, pad areas, and copper densities are weakened, making the overall defect severity index more focused on reflecting the deviation caused by the defect itself. The severity threshold can be set to... ,in Indicates the severity threshold. This represents the normalized fluctuation force of the good product impedance, and the coefficient 6 is used to reduce the probability of false rejection caused by good product fluctuation.

[0045] The judgment unit compares the comprehensive defect severity index with the severity threshold. When the current coordinate position meets the defect type judgment conditions and the comprehensive defect severity index reaches or exceeds the severity threshold, the system marks the PCBA board under test as a defective product and writes the defect coordinates, defect type and severity index into the detection record. At the same time, it outputs a rejection trigger signal to the online rejection mechanism. When only the angle condition is met but the severity index is insufficient, the system retains the abnormal record and continues to complete the full board scan without immediately triggering the rejection action, thereby reducing misjudgments caused by local noise, lifting residuals or slight process fluctuations.

[0046] This embodiment also provides a PCBA circuit board for implementing the steps of any of the methods described above, which will not be repeated here.

[0047] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0048] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for online performance testing of PCBA circuit boards, characterized in that, include: Excitation current is synchronously applied to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer and inner layer in the copper conductor of the PCBA board under test, and obtain the induced voltage signal at each probe position. The induced voltage signal is digitally phase-sensitive demodulated to obtain a set of impedance measurements at each probe position. A first impedance vector is constructed based on the impedance measurement set, and a second impedance vector is obtained by subtracting the first impedance vector from the reference impedance vector of the good product. The target phase angle is calculated based on the second impedance vector, the defect type is determined based on the target phase angle, and an online rejection decision is executed.

2. The online performance testing method for PCBA circuit boards according to claim 1, characterized in that, The process involves synchronously applying excitation current to the array probes to establish an eddy current induction field covering the surface layer, solder joint layer, and inner layer of the copper conductor in the PCBA board under test, thereby obtaining the induced voltage signal at each probe position, including: The array probes are scanned row by row and column coordinates, and an excitation current superimposed by three excitation frequencies is applied simultaneously. Based on the eddy current penetration depth corresponding to each excitation frequency, an eddy current induction field covering the surface layer, solder joint layer and inner layer is established in the copper conductor of the PCBA board under test. Based on the eddy current sensing field, the induced voltage signal at each probe position is collected.

3. The online performance testing method for PCBA circuit boards according to claim 2, characterized in that, The acquisition of induced voltage signals at each probe location based on the eddy current sensing field includes: Based on the electromagnetic induction between the eddy current induction field and the copper conductor of the PCBA board under test, an induction port voltage carrying information on the impedance change of the copper conductor is generated at the induction port of each probe coil. The voltage at the sensing port is input to an analog-to-digital converter for digital processing to obtain the induced voltage signal at each probe position.

4. The online performance testing method for PCBA circuit boards according to claim 1, characterized in that, The step of performing digital phase-sensitive demodulation on the induced voltage signal to obtain a set of impedance measurements for each probe position includes: The induced voltage signal is multiplied by the in-phase reference signal and the quadrature reference signal corresponding to each excitation frequency to obtain the in-phase filter component and the quadrature filter component of each excitation frequency. Within an integer multiple of the complete cycle of each excitation frequency, a discrete integration operation is performed on the in-phase filter component and the quadrature filter component to extract the impedance measurement set for each probe position.

5. The online performance testing method for PCBA circuit boards according to claim 4, characterized in that, The step of performing phase-sensitive multiplication operations on the induced voltage signal with the in-phase reference signal and quadrature reference signal corresponding to each excitation frequency to obtain the in-phase filter component and quadrature filter component of each excitation frequency includes: Based on the excitation frequencies output by the direct digital frequency synthesis module, in-phase reference signals and quadrature reference signals corresponding to the excitation current are generated in the digital domain. The induced voltage signal is multiplied point by point with the in-phase reference signal and the quadrature reference signal to obtain the in-phase multiplication integral and the quadrature multiplication integral corresponding to each excitation frequency; The in-phase multiplicative integral and the quadrature multiplicative integral are respectively filtered by a low-pass filter to remove the second harmonic component and the frequency aliasing term, so as to obtain the in-phase filtered component and the quadrature filtered component of each excitation frequency.

6. The online performance testing method for PCBA circuit boards according to claim 5, characterized in that, Within a complete cycle that is an integer multiple of each excitation frequency, a discrete integration operation is performed on the in-phase filter component and the quadrature filter component to extract the impedance measurement set for each probe position, including: Within the integration interval of an integer multiple of the excitation frequency, a discrete integration operation is performed on the in-phase filter component corresponding to each excitation frequency to obtain the real component of the first impedance corresponding to each excitation frequency. Within the integration interval of an integer multiple of the excitation frequency, a discrete integration operation is performed on the orthogonal filter components corresponding to each excitation frequency to obtain the first imaginary impedance component corresponding to each excitation frequency. The real component of the first impedance corresponding to each excitation frequency is combined with the imaginary component of the first impedance to obtain the impedance measurement set for each probe position.

7. The online performance testing method for PCBA circuit boards according to claim 6, characterized in that, The step of constructing a first impedance vector based on the impedance measurement set, and subtracting the first impedance vector from the reference impedance vector of the good product to obtain a second impedance vector, includes: Based on the preset lift-off sensitivity vector, the first real component and the first imaginary component of the impedance corresponding to each excitation frequency in the impedance measurement group are lifted off and compensated respectively to obtain the second real component and the second imaginary component of the impedance. The real component of the second impedance is combined with the imaginary component of the second impedance to form a first impedance vector; The first impedance vector is subtracted from the reference impedance vector of the good product to obtain the real part difference component and the imaginary part difference component of the impedance corresponding to each excitation frequency. The real part difference component and the imaginary part difference component of the impedance are then combined to form the second impedance vector.

8. The online performance testing method for PCBA circuit boards according to claim 1, characterized in that, The step of calculating the target phase angle based on the second impedance vector, determining the defect type based on the target phase angle, and performing an online rejection decision includes: On the complex impedance plane corresponding to each excitation frequency, the four-quadrant arctangent operation is performed on the difference components of the real part of the impedance and the difference components of the imaginary part of the impedance corresponding to each excitation frequency in the second impedance vector to obtain the target phase angle corresponding to each excitation frequency. The defect type is determined based on the target phase angle, and the comprehensive defect severity index is calculated using the reference impedance mode corresponding to each excitation frequency. An online rejection decision is then made for the PCBA board under test based on the comprehensive defect severity index.

9. The online performance testing method for PCBA circuit boards according to claim 8, characterized in that, The process of determining the defect type based on the target phase angle, calculating the comprehensive defect severity index using the reference impedance modulus corresponding to each excitation frequency, and performing an online rejection decision on the PCBA board under test based on the comprehensive defect severity index includes: The defect type is determined based on whether the amplitude of the target phase angle corresponding to each excitation frequency and the amplitude of the corresponding impedance real part difference component exceed the corresponding amplitude threshold. Based on the real part difference component and the imaginary part difference component of the impedance, the first difference component magnitude and the second difference component magnitude of the real part difference component of the impedance corresponding to each excitation frequency in the second impedance vector are calculated respectively. A weighted summation is performed on the first difference component magnitude and the second difference component magnitude to obtain the comprehensive defect severity index; The comprehensive defect severity index is compared with the severity threshold; If the overall defect severity index exceeds the severity threshold, an online rejection decision is triggered for the PCBA board under test.

10. A PCBA circuit board, characterized in that, The steps are for implementing the online performance testing method for PCBA circuit boards according to any one of claims 1 to 9.

Images