Circuit board conductivity rapid detection device

Abstract

The invention relates to the technical field of electronic product quality detection, and discloses a circuit board conductivity rapid detection device which comprises a multi-band millimeter wave cooperative scanning module, a phased-array antenna module, a three-dimensional positioning module, a four-dimensional defect evolution model module and a signal processing module. The device performs nondestructive testing through a multi-band millimeter wave technology, realizes self-adaptive beam forming by combining a phased-array antenna, positions internal conductive defects by adopting multi-angle scanning and a three-dimensional reconstruction algorithm, constructs a four-dimensional defect evolution model to track a defect expansion trend, and evaluates the conductivity through a special signal processing algorithm. The technical problems that a traditional detection method is low in efficiency and insufficient in precision, and dynamic monitoring cannot be carried out are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of quality inspection of electronic products, and more specifically, it relates to a device for rapidly detecting the electrical conductivity of a circuit board. Background Art

[0002] With the development of miniaturization and high integration of electronic devices, the structure of circuit boards has become increasingly complex, and their reliability and electrical conductivity have become key factors affecting the quality of the whole machine. Currently, the detection of the electrical conductivity of circuit boards mainly uses technical means such as needle testing, X-ray detection, and contact electrical testing, but these traditional methods have obvious deficiencies. The needle testing method requires disassembling or damaging the sample to detect the inner-layer electrical conductivity, with low efficiency and easy to cause damage to the circuit board; although X-ray detection is a non-destructive detection, the detection rate for tiny electrical defects is not high, and the detection speed is slow; contact electrical testing is limited by the probe density and is difficult to meet the detection requirements of high-density tiny circuits.

[0003] More critically, the existing detection technologies mainly provide static detection results and cannot monitor the dynamic evolution process of electrical defects under actual working environments (such as thermal cycling, mechanical vibration), resulting in a large number of potential faults unable to be warned in time. At the same time, when the defect is located in a complex laminated structure, the positioning accuracy of traditional detection means is insufficient, making it difficult to accurately indicate the fault location, increasing the maintenance difficulty and cost. In addition, the speed of traditional detection methods is difficult to meet the rapid detection requirements of mass production, becoming a bottleneck in production efficiency.

[0004] Therefore, there is an urgent need for a new detection system that can detect the electrical conductivity of multi-layer circuit boards non-destructively, quickly, and accurately, and can realize dynamic monitoring of defects to solve the above technical problems. Summary of the Invention

[0005] The present invention provides a device for rapidly detecting the electrical conductivity of a circuit board, which solves the technical problems of low detection efficiency, insufficient accuracy, and inability to perform dynamic monitoring in related technologies.

[0006] The present invention provides a device for rapidly detecting the electrical conductivity of a circuit board, including: A multi-band millimeter-wave collaborative scanning module, including a millimeter-wave transmitting unit, a receiving unit, a signal processing unit, and a control unit, for realizing selective detection of different layers of the circuit board; A phased array antenna module, for realizing adaptive beamforming, so that the millimeter wave can focus on a specific area of the circuit board for key monitoring; A three-dimensional positioning module, for performing multi-angle scanning and applying a three-dimensional reconstruction algorithm to realize precise spatial positioning of internal electrical defects; A four-dimensional defect evolution model module, for obtaining data on the change of defects over time through repeated scanning and real-time tracking of the expansion trend of inner-layer defects; A signal processing module, which is used to apply a reflection characteristic analysis algorithm to determine the integrity and conductivity of a conductive path.

[0007] Furthermore, the multi-band millimeter-wave collaborative scanning module includes millimeter-wave transmitters in three frequency bands of 30 - 40 GHz, 60 - 70 GHz, and 75 - 85 GHz. Among them, the 30 - 40 GHz frequency band is mainly used for deeper layer detection, the 60 - 70 GHz frequency band is used for intermediate layer detection, and the 75 - 85 GHz frequency band is used for surface and shallow layer detection.

[0008] Furthermore, in the phased array antenna module, the phase distribution of each antenna element satisfies the formula: ; where represents the phase of the th antenna element, is the initial phase, is the distance from this element to the reference point, is the millimeter-wave wavelength, is the beam pointing angle.

[0009] Furthermore, the operation of performing multi-angle scanning in the three-dimensional positioning module includes scanning the circuit board from a total of 12 angles in the positive and negative directions of the three main axes of X, Y, and Z, and collecting reflection data in three frequency bands at each scanning angle.

[0010] Furthermore, the function for realizing the precise spatial positioning of internal conductive defects by applying a three-dimensional reconstruction algorithm is: ; where represents the possibility of a defect existing at the spatial point , represents the reflection intensity of the rd frequency band at the th angle, is the corresponding weight coefficient, represents the number of frequency bands, represents the number of scanning angles.

[0011] Furthermore, in the four-dimensional defect evolution model module, the four-dimensional defect evolution model is based on the following mathematical formula: ; where represents the defect characteristic function of the spatial point at time , represents the defect characteristic of the initial state, represents the time The evolution velocity field at a moment.

[0012] Furthermore, the implementation of the four-dimensional defect evolution model includes discretizing the integral: ; Among them, represents the defect characteristic function of the spatial point at time moment, , respectively represent the , th time points, represents the interval between adjacent time points.

[0013] Furthermore, in the signal processing module, the deviation degree of the reflection characteristic is quantified by the following formula: ; Among them, represents the deviation between the measured reflection coefficient and the reference reflection coefficient, represents the measured reflection coefficient at frequency , represents the reference reflection coefficient.

[0014] Furthermore, it also includes constructing a signal-to-noise ratio enhancement algorithm module. The signal-to-noise ratio enhancement algorithm adopts a method combining adaptive filtering and signal accumulation, and enhances the signal strength through the following formula: ; Among them, is the enhanced signal spectrum, is the signal spectrum of the th scan, is the adaptive weight coefficient, is the number of scans.

[0015] The present invention provides a storage medium that stores non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are executed by a computer, the modules in the above-mentioned fast detection device for the electrical conductivity of a circuit board can be executed.

[0016] The beneficial effects of the present invention are as follows: The fast detection system for the electrical conductivity of the circuit board of the present invention solves the problems of low efficiency, poor accuracy, and inability to dynamically monitor in traditional detection methods, and has many advantages. It uses millimeter-wave technology to carry out non-destructive detection, avoiding the disassembly of samples and ensuring that the circuit board is intact and can be repeatedly detected; the scanning speed reaches 20 cm² per second, improving the detection efficiency; by using multi-band collaborative scanning and three-dimensional reconstruction algorithms, it can detect the smallest conductive path of 50 microns, and the detection rate of abnormal electrical conductivity is 99.2%; with the help of multi-angle scanning and three-dimensional reconstruction technology, three-dimensional positioning with an accuracy of ±0.1 mm for the inner-layer conductive defects of multi-layer circuit boards is achieved; through a four-dimensional defect evolution model, the dynamic evolution process of defects is monitored; and it is applicable to various complex circuit boards, providing a non-destructive, fast, and accurate quality detection solution for them, improving the reliability and service life of electronic products. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a module diagram of a fast detection device for the electrical conductivity of a circuit board of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and changes can be made to the functions and arrangements of the elements discussed without departing from the scope of protection of the content of this specification. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described in some examples can also be combined in other examples.

[0019] In at least one embodiment of the present invention, a fast detection device for the electrical conductivity of a circuit board is disclosed, as Figure 1 shown, including: A multi-band millimeter-wave collaborative scanning module, including a millimeter-wave transmitting unit, a receiving unit, a signal processing unit, and a control unit, for realizing selective detection of different layers of the circuit board; Specifically including: Step 1.1, configuring a multi-band millimeter-wave transmitter, which can simultaneously or alternately transmit millimeter-wave signals of three frequency bands, and using the penetration depth differences of millimeter-waves of different frequency bands to realize selective detection of different layers of the circuit board; Step 1.2, configuring a millimeter-wave receiving unit, including a multi-band antenna array and a signal receiving circuit, for receiving millimeter-wave signals of different frequency bands reflected from the circuit board; Step 1.3, setting a control unit, which is responsible for adjusting the transmission power, scanning angle, frequency selection, and scanning path planning to ensure full coverage of the entire circuit board; Step 1.4, establish the mapping relationship between millimeter waves of different frequency bands and the depth of the circuit board layer. The 30 - 40 GHz frequency band is mainly used for deeper layer detection, the 60 - 70 GHz frequency band is used for intermediate layer detection, and the 75 - 85 GHz frequency band is used for surface and shallow layer detection, forming a multi-depth collaborative detection system.

[0020] The phased array antenna module is used to achieve adaptive beamforming, enabling the millimeter wave to focus on a specific area of the circuit board for key monitoring; Specifically include: Step 2.1, construct the phased array antenna module. By controlling the phase of each antenna element, the dynamic adjustment of the beam is realized. The phase distribution of each antenna element satisfies the formula: ; Where, represents the phase of the i-th antenna element, is the initial phase, is the distance from this element to the reference point, is the wavelength of the millimeter wave, is the beam pointing angle; Step 2.2, construct an adaptive control algorithm. According to the preliminary scan results, automatically identify the areas that need to be key detected and dynamically adjust the beam direction and focusing degree; this algorithm is based on the following mathematical model: ; Where, represents the anomaly degree score at position , represents the reflection anomaly value of the -th frequency band at this position, is the weight coefficient of the corresponding frequency band, is the total number of frequency bands. When exceeds the preset threshold , this area is marked as the key detection area.

[0021] Step 2.3, achieve multi-beam parallel processing. By grouping the phased array antennas, multiple detection beams are formed simultaneously to improve the scanning efficiency; the multi-beam formation satisfies the formula: ; Where, represents the gain of the -th beam in the direction , is the amplitude of the -th antenna element in the -th group, is the phase, is the element position, is the number of antenna elements in this group, is the working wavelength. , is the beam direction.

[0022] Step 2.4: Establish a feedback optimization mechanism to continuously optimize beam parameters and improve detection accuracy based on the real-time received signal quality. The optimization process uses an iterative method: ; in, Indicates The iteration Group The phase of the antenna elements, Indicates The iteration Group The phase of the antenna elements, is the signal quality evaluation function, Its gradient, is the learning rate.

[0023] The signal quality evaluation function is defined as: ; in, The received power for the target area, is the total transmit power, is the signal-to-noise ratio.

[0024] A 3D positioning module that performs multi-angle scanning and applies a 3D reconstruction algorithm to achieve precise spatial positioning of internal conductive defects; Specifically include: Step 3.1, perform multi-angle scanning, perform millimeter wave scanning at at least 12 different angles around the circuit board, and each angle covers the entire circuit board; Step 3.2, apply the 3D reconstruction algorithm to construct the defect location function based on the reflection data obtained from different angles: ; in, Representing a spatial point The possibility of defects, represents the reflection intensity at the jth angle in the i-th frequency band, is the corresponding weight coefficient, m represents the number of frequency bands, and n represents the number of scanning angles; The 3D reconstruction algorithm is based on the principle of reverse projection. The specific steps are: first, the 2D scan data of each angle is mapped to the 3D space through reverse projection transformation, and then the 3D data of different angles are weighted and fused. The determination of is done using an adaptive method based on the signal quality evaluation function: ; Among them, represents the signal-to-noise ratio at the j-th angle in the i-th frequency band.

[0025] In practical applications, taking the detection of micro-cracks in the internal conductive paths of a four-layer PCB board as an example, the module first performs scans at 4 angles in each of the three main axes of X, Y, and Z, for a total of 12 angles. At each angle, millimeter waves in three frequency bands (30 - 40 GHz, 60 - 70 GHz, 75 - 85 GHz) are used. The 36 sets of two-dimensional data obtained are fused through a three-dimensional reconstruction algorithm, and finally a three-dimensional defect distribution map with a spatial resolution of 0.1 mm is obtained; Step 3.3, use an iterative optimization algorithm to gradually refine the reconstruction result and improve the clarity of the defect boundary; this algorithm is based on the following mathematical model: ; Among them, represents the defect distribution function after the -th iteration, represents the defect distribution function of the -th iteration, represents the Laplace operator, which is used to extract boundary information, represents the regularization function, which is used to suppress noise, is the iteration step parameter, which controls the convergence speed, is the regularization parameter, which is used to balance the boundary clarity and noise suppression. The iteration termination condition is: ; Among them, is the preset convergence threshold, usually taken as 0.001; represents the defect distribution function after the -th iteration, represents the defect distribution function of the -th iteration.

[0026] Step 3.4, establish a defect feature library, and identify the defect type (such as open circuit, short circuit, crack) by comparing the reconstruction result with the pre-established defect feature library. The feature matching degree calculation formula is: ; Among them, represents the defect to be detected and the -th defect template in the feature library The matching degree ranges from , and the larger the value, the higher the matching degree. The module selects the one with the highest matching degree and exceeding the threshold The defect type with a (usually taken as 0.85) is used as the recognition result. If all the matching degrees are lower than the threshold, it is marked as a defect of unknown type. represents the distribution function of the defect to be detected, represents the th class defect template in the feature library.

[0027] Four-dimensional defect evolution model module, used to obtain data on the change of defects over time through repeated scanning, and to track the expansion trend of inner-layer defects in real time; Specifically including: Step 4.1, construct a dynamic monitoring module, and obtain data on the change of defects over time by repeatedly performing the aforementioned scanning at specific time intervals; Step 4.2, construct a four-dimensional defect evolution model, and this model is based on the following mathematical formula: ; where, represents the defect feature function of the spatial point (x, y, z) at time t, represents the defect feature in the initial state, represents the evolution velocity field of this point at time ; In specific implementation, the integral is discretized: ; where, , respectively represent the , th time points, represents the interval between adjacent time points.

[0028] In practical applications, taking the expansion of micro-cracks on a PCB board as an example, first obtain the three-dimensional distribution of the cracks in the initial state , then perform a scan every 10 minutes under temperature cycling conditions (such as -40°C to 85°C, with a cycle period of 1 hour), calculate the expansion speed of the cracks , and finally obtain the crack distribution state at different time points ; Step 4.3, apply a trend prediction algorithm, and based on historical evolution data, predict the development trend of the defect at future time points, providing a basis for preventive maintenance; The trend prediction algorithm uses a method combining polynomial fitting and exponential smoothing to model the evolution process of the defect. The algorithm first analyzes the change sequence of the defect feature values of each spatial point over time, where, , , respectively represent the , , time points, , , are respectively the , , defect characteristic values of the spatial points at the time points, and identify their change patterns. For the linear growth pattern, linear regression prediction is adopted; for the accelerated expansion pattern, quadratic or exponential function fitting is adopted; for the periodic change pattern, Fourier analysis is used to extract the periodic characteristics and then prediction is carried out.

[0029] In practical applications, taking the microcracks in a high-density interconnect board (HDI) as an example, the module collects the expansion data of the cracks at 20 time points under temperature cycling conditions and identifies that the crack length shows an exponential growth pattern , where is the expansion coefficient. Through this model, the module predicts that the crack will reach the critical length after 72 hours and issues a replacement warning to the maintenance personnel in advance; Step 4.4, construct an environmental simulation unit to simulate the actual working environment of thermal cycling and mechanical vibration, accelerate the defect evolution process, and discover potential failure risks in advance.

[0030] The environmental simulation unit is based on the following mathematical model: ; where represents the defect area after time at temperature , represents the initial defect area, represents the temperature acceleration coefficient, represents the reference temperature, represents time.

[0031] For the mechanical vibration acceleration model, the following formula is adopted: ; where represents the defect length after time under the vibration condition with frequency and acceleration , represents the initial defect length, represents the vibration sensitivity coefficient, represents the frequency exponent (usually taking values of 1.5 - 2.0), represents the acceleration exponent (usually taking values of 1.0 - 1.3), represents time.

[0032] The comprehensive acceleration model calculates the equivalent working time through the following formula: ; Wherein, represents the equivalent working time, represents the test time, represents the total acceleration factor, represents the temperature acceleration factor, represents the vibration acceleration factor. The calculation formula for the temperature acceleration factor is: , where is the activation energy, is the Boltzmann constant, is the actual operating temperature, is the test temperature. The calculation formula for the vibration acceleration factor is: , where is the test acceleration, is the actual operating acceleration.

[0033] A signal processing module, which is used to apply the reflection characteristic analysis algorithm to determine the integrity and conductivity of the conductive path; Specifically includes: Step 5.1, applying the reflection characteristic analysis algorithm, and determining the integrity of the conductive path by comparing and analyzing the millimeter-wave reflection characteristics of different conductive materials and states; The reflection characteristic analysis algorithm is optimized for the detection of the conductivity of the circuit board based on the electromagnetic wave scattering theory. The algorithm first establishes a reference reflection model according to the electromagnetic parameters of the circuit material, and then compares the actually measured reflection characteristics with the reference model to quantify the deviation degree. The specific mathematical expression is: ; Wherein, represents the deviation between the actually measured reflection coefficient and the reference reflection coefficient, represents the frequency of the actually measured reflection coefficient, represents the reference reflection coefficient.

[0034] The algorithm identifies the type of conductive anomaly by analyzing the distribution characteristics of at different frequencies. For example, an open circuit usually shows a significant increase in the reflection coefficient in a specific frequency band, while a microcrack shows a small increase in the high-frequency reflection coefficient.

[0035] In practical applications, taking the detection of the virtual solder joints on the circuit board as an example, the module first scans the standard solder joints to establish a reference reflection model, then scans the solder joints to be tested and applies the reflection characteristic analysis algorithm, and finds that the reflection characteristics in the 65 - 70 GHz frequency band deviate from the normal value by more than 20%, and accordingly judges it as a virtual solder joint; Step 5.2, construct a signal-to-noise ratio enhancement algorithm. By means of multiple sampling and signal filtering techniques, improve the detection ability of weak signals, enabling the module to detect subtle conductive anomalies; The signal-to-noise ratio enhancement algorithm adopts a method combining adaptive filtering and signal accumulation. First, the received signal is subjected to wavelet transform and decomposed into signal components of different scales; then, according to the signal-to-noise ratio characteristics of each component, adaptive Wiener filtering is selectively applied; finally, the results of multiple scans are coherently accumulated to increase the signal strength. The core mathematical formula of the algorithm is: ; where, is the enhanced signal spectrum, is the signal spectrum of the th scan, is the adaptive weight coefficient, and K is the number of scans.

[0036] The weight coefficient is determined through signal-to-noise ratio estimation: ; where, is the signal-to-noise ratio estimation value at the kth scan at frequency .

[0037] In practical applications, for the conductivity detection of high-density micro-lines (line width 50 μm), the module processes through the signal-to-noise ratio enhancement algorithm, increasing the detection sensitivity by 3.5 dB and successfully identifying micro-cracks that cannot be detected by traditional methods; Step 5.3, implement quantitative evaluation of conductivity. Based on the amplitude, phase, and frequency response characteristics of the reflected signal, establish a quantitative index system for conductivity; the quantitative evaluation of conductivity adopts the comprehensive index method, and calculates the conductivity index through the following mathematical model: ; where, represents the conductivity performance index (ConductivityPerformanceIndex), with a value range of 0 - 100, and the higher the value, the better the conductivity; represents the weight coefficient of the ith parameter, satisfying ; represents the measured value of the ith parameter; and respectively represent the minimum allowable value and the maximum allowable value of the ith parameter. The key parameters include: Reflection signal amplitude deviation rate: ; where, and respectively represent the measured amplitude and the reference amplitude at the frequency , and m represents the number of sampling frequency points.

[0038] Phase consistency index: ; where and respectively represent the measured phase and the reference phase at the frequency .

[0039] Frequency response correlation coefficient: ; where represents the covariance, represents the standard deviation, and respectively represent the measured frequency response and the reference frequency response.

[0040] Conductive path continuity index: ; where represents the number of detected discontinuity points, represents the total number of detected points on the conductive path.

[0041] Step 5.4, construct a hierarchical alarm mechanism. According to the evaluation results of the electrical conductivity, classify the anomalies into different levels to provide a priority reference for maintenance and handling. The hierarchical alarm mechanism is based on the electrical conductivity index and the defect growth rate (GrowthRate) to construct a risk matrix model: ; where represents the risk level (RiskLevel), and respectively represent the weight coefficients of the electrical conductivity index and the defect growth rate. Usually . The defect growth rate The calculation formula is: ; where represents the defect characteristic function of the spatial point at time .

[0042] According to the risk level , classify the anomalies into four levels: Level 1 alarm (urgent): , indicating that there is a serious electrical conductivity anomaly and immediate shutdown for maintenance is required; Secondary alarm (high risk): It indicates obvious conductive abnormality and requires maintenance within 24 hours; Tertiary alarm (medium risk): It indicates slight conductive abnormality and requires maintenance within one week; Quaternary alarm (low risk): It indicates potential conductive abnormality and requires regular monitoring.

[0043] The following technical effects are achieved in this embodiment: Non-destructive testing effect: The millimeter-wave technology is adopted to achieve non-destructive testing of the circuit board, avoiding the disadvantages of traditional methods that require disassembling or damaging samples, enabling comprehensive testing of the complete circuit board, and ensuring the integrity and repeatable detectability of the circuit board.

[0044] High-efficiency testing: The scanning speed of the module can reach 20 cm² per second, greatly improving the testing efficiency, making quality control in mass production no longer a bottleneck for production efficiency, and the testing speed is 5 to 10 times higher than that of traditional methods.

[0045] High-precision testing: Through multi-band collaborative scanning and 3D reconstruction algorithms, the module can detect conductive paths as small as 50 microns, and the detection rate of abnormal conductivity reaches 99.2%, significantly superior to traditional detection methods.

[0046] Precise positioning of inner-layer defects: The multi-angle scanning and 3D reconstruction technology realizes precise 3D positioning of conductive defects inside multi-layer circuit boards, with a positioning accuracy of ±0.1 mm, providing accurate position information for subsequent maintenance.

[0047] Dynamic evolution monitoring ability: The four-dimensional defect evolution model enables the system to have the ability to monitor the dynamic evolution process of defects in the actual working environment, developing the detection of the inner-layer conductive performance of circuit boards from static photographing to dynamic video recording, providing technical support for preventive maintenance.

[0048] Wide range of applications: The system is applicable to various circuit boards, especially high-density multi-layer circuit boards, complex circuit boards used in high-end servers and communication equipment, providing a non-destructive, fast and accurate quality detection method, and greatly improving the reliability and service life of electronic products.

[0049] In an embodiment of the present invention, an example of the foregoing rapid detection system for the conductive performance of circuit boards is provided; Description of application scenario: This embodiment is applied to the production line of high-density multi-layer PCB boards of a certain communication equipment manufacturing enterprise, mainly solving the following problems encountered by the enterprise in the production process: Traditional needle detection methods cannot detect high-density inner-layer conductive anomalies; The production line speed requires the detection of 300 PCB boards per hour, while the existing methods can only reach 80 boards per hour; Inner-layer defects cannot be accurately located, resulting in difficult repair; 30% of the customer repair reports are caused by the deterioration of potential defects during use.

[0050] The application scenario is specifically the quality inspection of 8-layer HDI (High-Density Interconnect) boards. The size of the HDI board is 25cm×18cm, the minimum line width is 50μm, and there are a large number of micro-vias and blind-via interconnection structures in the inner layer, which are used in 5G base station equipment.

[0051] Implementation process example: Implementation of the multi-band millimeter-wave collaborative scanning system: The enterprise constructed a multi-band millimeter-wave collaborative scanning system according to this implementation method and configured the following hardware devices, as shown in Table 1: Table 1: Hardware configuration of the multi-band millimeter-wave collaborative scanning system:

[0052] After the system was installed, calibration tests were carried out on standard test boards, and the mapping relationship between millimeter waves of different bands and the hierarchical depth of the PCB board was established. It was determined that the 30-40 GHz band was mainly used to detect the 5th - 8th layers, 60-70 GHz was used to detect the 3rd - 4th layers, and 75-85 GHz was used to detect the 1st - 2nd layers.

[0053] Implementation of adaptive beamforming: For the key monitoring of high-density areas (especially micro-via and BGA pad areas) in HDI boards, the system implemented the adaptive beamforming function. In actual applications, the system first performs a low-resolution scan of the entire board. After identifying suspicious areas, it automatically adjusts the beam parameters to focus on these areas for high-resolution scanning, as shown in Table 2: Table 2: Adaptive beamforming parameter settings:

[0054] During the detection of the BGA pad area of the HDI board, the system first found abnormal reflection characteristics in a certain area through global scanning, then automatically switched to the area focusing scan mode, further found three suspicious pads, and finally used fine focusing scanning to confirm that there was a small crack in one of the pads. The entire process took only 2.5 seconds.

[0055] Implementation of multi-angle scanning and three-dimensional defect location: In the actual application of this enterprise, multi-angle scanning of each HDI board was carried out at 12 angles, including 2 angles in each positive and negative direction of the X, Y, and Z axes. Reflection data in three frequency bands were collected at each scanning angle, and data fusion and defect location were performed through a three-dimensional reconstruction algorithm, as shown in Table 3: Table 3: Multi-angle scanning configuration:

[0056] In a certain detection, the system successfully located a break point of the conductive path on the 6th layer of the HDI board through a three-dimensional reconstruction algorithm. This break point is between two micro-vias and has a width of only 60μm, which cannot be detected by traditional detection methods. The three-dimensional coordinates output by the system are (127.53mm, 86.21mm, 1.25mm), accurately indicating the defect location. The maintenance personnel successfully located and repaired the fault based on this coordinate.

[0057] Implementation of the four-dimensional defect evolution model: To solve the problem of failures caused by the deterioration of potential defects during the customer's use process, this enterprise established a defect evolution model laboratory and conducted sampling accelerated aging tests on mass-produced PCB boards to simulate the defect evolution process in the actual use environment, as shown in Table 4: Table 4: Defect evolution test conditions:

[0058] Through sampling tests on 10 batches of a total of 100 HDI boards, the system constructed a four-dimensional defect evolution model and successfully predicted the evolution trends of three common types of defects (micro-cracks, solder joint voids, and stress damage to inner-layer conductors). For the identified potential risk PCB boards, the system will give the expected failure time and recommended treatment measures.

[0059] In a certain batch of products, the system found that there were small cracks near the micro-vias on 5 PCB boards. Although the current electrical conductivity was normal, through the prediction of the four-dimensional defect evolution model, these cracks would expand to the extent of causing open circuits in about 6 months. Based on this, the enterprise replaced these PCB boards in advance and avoided the occurrence of client-side failures.

[0060] Verification of technical effects: Through the actual application in this communication equipment manufacturing enterprise, two core technical effects of this implementation method were verified: improved detection efficiency and increased defect detection rate.

[0061] Verification of improved detection efficiency, as shown in Table 5: Table 5: Comparison of detection efficiency:

[0062] It can be seen that by adopting the multi-band millimeter wave technology of this embodiment, the detection time of a single HDI board is only 12 seconds, which is 225 times higher than that of the traditional needle detection method and still 37.5 times higher than that of the most advanced contact electrical testing method, meeting the detection requirement of 300 PCB boards per hour on the production line.

[0063] Verification of the improvement in the defect detection rate is shown in Table 6: Table 6: Comparison of the detection rates of different types of defects:

[0064] It can be seen that the test results for 200 known defect samples show that the overall defect detection rate of this embodiment reaches 95.8%, which is 43.8 percentage points higher than that of the traditional method. In particular, the detection ability for inner layer defects has been significantly improved. The detection rate of inner layer microcracks has increased from 36% to 96%, and the detection rate of inner layer conductor stress damage has increased from 20% to 92%.

[0065] One year after implementing this method in this enterprise, the defective rate of products leaving the factory has decreased from 2.8% to 0.3%, the failure rate at the client side has decreased from 1.5% to 0.2%, the maintenance cost has been reduced by 78%, and the product reliability and customer satisfaction have been significantly improved.

[0066] The embodiments of the present invention have been described above. However, these embodiments are not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative and not restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make more equivalent embodiments in various forms, all of which fall within the protection scope of this embodiment.

Claims

1. A rapid detection device for the electrical conductivity of a circuit board, characterized in that, Including: A multi-band millimeter-wave collaborative scanning module, including a millimeter-wave transmitting unit, a receiving unit, a signal processing unit, and a control unit, for realizing selective detection of different layers of a circuit board; A phased array antenna module, for realizing adaptive beamforming, enabling the millimeter-wave to focus on a specific area of the circuit board for key monitoring; A three-dimensional positioning module, for performing multi-angle scanning and applying a three-dimensional reconstruction algorithm to achieve precise spatial positioning of internal conductive defects; A four-dimensional defect evolution model module, for obtaining data on the change of defects over time through repeated scanning, and real-time tracking the expansion trend of inner-layer defects; A signal processing module, for applying a reflection characteristic analysis algorithm to determine the integrity and conductivity of the conduction path.

2. A fast detection device for the electrical conductivity of a circuit board according to claim 1, characterized in that, The multi-band millimeter-wave collaborative scanning module includes millimeter-wave transmitters in three frequency bands of 30-40 GHz, 60-70 GHz, and 75-85 GHz. Among them, the 30-40 GHz frequency band is mainly used for deeper layer detection, the 60-70 GHz frequency band is used for intermediate layer detection, and the 75-85 GHz frequency band is used for surface and shallow layer detection.

3. A rapid detection device for the electrical conductivity of a circuit board according to claim 1, characterized in that, In the phased array antenna module, the phase distribution of each antenna element satisfies the formula: ； Among them, represents the phase of the th antenna element, is the initial phase, is the distance from this element to the reference point, is the millimeter-wave wavelength, is the beam pointing angle.

4. A rapid detection device for the electrical conductivity of a circuit board according to claim 1, characterized in that, The operation of performing multi-angle scanning in the three-dimensional positioning module includes scanning the circuit board from a total of 12 angles in the positive and negative directions of the three main axes of X, Y, and Z, and collecting reflection data of three frequency bands at each scanning angle.

5. The rapid detection device for the conductive performance of a circuit board according to claim 1, wherein The function of applying a three-dimensional reconstruction algorithm to achieve precise spatial positioning of internal conductive defects is: ； in, Representing a spatial point The possibility of defects, Indicates Frequency Band The reflection intensity of the angle, is the corresponding weight coefficient, Indicates the number of frequency bands, Indicates the number of scan angles.

6. The rapid detection device for the electrical conductivity of a circuit board according to claim 1, wherein In the four-dimensional defect evolution model module, the four-dimensional defect evolution model is based on the following mathematical formula: ； Among them, represents the spatial point at time the defect feature function at the moment, represents the defect feature in the initial state, represents the evolution velocity field of this point at time at the moment.

7. A rapid detection device for the electrical conductivity of a circuit board according to claim 6, characterized in that, The implementation of the four-dimensional defect evolution model includes discretizing the integral: ； Among them, represents the spatial point at time the defect feature function at the moment, and respectively represent the and th time points, represents the interval between adjacent time points.

8. A rapid detection device for the electrical conductivity of a circuit board according to claim 1, characterized in that, In the signal processing module, the deviation degree of the reflection characteristic is quantified by the following formula: ； wherein, represents the deviation between the measured reflection coefficient and the reference reflection coefficient, represents the frequency of the measured reflection coefficient, and represents the reference reflection coefficient.

9. A rapid detection device for the electrical conductivity of a circuit board according to claim 1, characterized in that, It also includes constructing a signal-to-noise ratio enhancement algorithm module. The signal-to-noise ratio enhancement algorithm uses a method combining adaptive filtering and signal accumulation to enhance the signal strength by the following formula: ； Among them, is the enhanced signal spectrum, is the signal spectrum of the th scan, is the adaptive weight coefficient, is the number of scans.

10. A storage medium stores non - transitory computer - readable instructions, characterized in that, When the non-transitory computer-readable instructions are executed by a computer, they can execute the modules in a fast circuit board conductivity detection device as described in any one of claims 1-9.

Images