Solder paste double-printing method and device for PCBA surface mounting technology

Abstract

The invention relates to the technical field of surface mounting processes, and discloses a solder paste double-printing method and equipment for a PCBA (Printed Circuit Board Assembly) surface mounting process. The method comprises the following steps: scanning the surface of a PCBA (Printed Circuit Board Assembly) subjected to first solder paste printing, obtaining solder paste morphology data and extracting actual coordinates of a BGA (Ball Grid Array) bonding pad from the solder paste morphology data; constructing a coordinate transformation matrix of the actual coordinates and the design coordinates, and creating a snakelike printing path of the second solder paste printing according to the coordinate transformation matrix; solder paste state parameters on the surface of the PCBA are detected, and when the solder paste state parameters meet printing conditions, a starting signal is generated; and the starting signal is received, second-time solder paste printing is executed according to the S-shaped printing path, real-time compensation is carried out, and the printed PCBA is obtained. According to the invention, the influence of PCB manufacturing tolerance and device mounting deviation on the printing precision is effectively compensated, and the alignment precision and consistency of double printing are improved.

Description

Technical Field

[0001] This invention relates to the field of surface mount technology, and in particular to a method and equipment for dual solder paste printing in PCBA surface mount technology. Background Technology

[0002] Due to the unique structure of BGA (Ball Grid Array) devices, the thickness and precision requirements of solder paste printing are extremely high. Traditional single-pass printing processes often fail to meet the solder paste volume requirements of BGA pads, leading to quality issues such as cold solder joints and poor solderability after soldering. To solve this problem, a dual-printing process is currently commonly used. After the first full-board solder paste printing is completed, a selective second printing is performed on the BGA area to ensure that the BGA pads receive sufficient solder paste volume.

[0003] However, due to factors such as PCB manufacturing tolerances, component placement deviations, and solder paste deformation after the first printing, the actual position of BGA pads in existing dual printing technology often deviates from the design position. This leads to problems such as low printing recognition accuracy and inaccurate boundary division, ultimately affecting the overall soldering quality and reliability of the board. Summary of the Invention

[0004] The main objective of this invention is to provide a method and equipment for double solder paste printing in PCBA surface mount technology. This invention effectively compensates for the impact of PCB manufacturing tolerances and component placement deviations on printing accuracy, and improves the alignment accuracy and consistency of double printing.

[0005] To achieve the above objectives, the present invention provides a method for dual solder paste printing in PCBA surface mount technology, comprising the following steps: Scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data, and extract the actual coordinates of the BGA pads from it; Construct a coordinate transformation matrix between the actual coordinates and the design coordinates, and create a serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; The solder paste state parameters on the PCBA surface are detected, and a start signal is generated when the solder paste state parameters meet the printing conditions. Upon receiving the start signal, the second solder paste printing is performed according to the serpentine printing path, and real-time compensation is performed to obtain the printed PCBA.

[0006] Optionally, in a first implementation of the first aspect of the present invention, the step of scanning the PCBA surface after the first solder paste printing is completed, obtaining solder paste morphology data, and extracting the actual coordinates of the BGA pads from it includes: The laser is activated to measure the height of the PCBA surface after the first solder paste printing is completed, and three-dimensional coordinate data is obtained. The three-dimensional coordinate data is subjected to Gaussian filtering to form height measurement data; The exposed area of ​​the PCBA substrate in the height measurement data is selected as the reference plane, and the height difference of each measurement point relative to the reference plane is calculated to generate a topographic image. The layout information in the PCBA design file is matched and marked with the topographic image to identify the expected location area of ​​BGA devices, the pad area of ​​ordinary devices and the blank area of ​​the substrate, and integrated into solder paste topographic data. The actual coordinates of the BGA pads are extracted from the solder paste morphology data.

[0007] Optionally, in a second implementation of the first aspect of the present invention, the step of extracting the actual coordinates of the BGA pads from the solder paste topography data includes: Using a pre-set thickness threshold, high-thickness pixels exceeding the thickness threshold are selected from the solder paste morphology data to form a candidate pad distribution map. The candidate pad distribution map is opened and closed using a circular structuring element to obtain the effective pad outline. The spacing between the effective pad outlines is measured and matched with the standard BGA ball spacing to confirm the BGA pad array, and the actual coordinates of each BGA pad in the BGA pad array are calculated.

[0008] Optionally, in a third implementation of the first aspect of the present invention, the step of measuring the spacing between the effective pad outlines and matching it with the standard BGA ball pitch to confirm the BGA pad array, and calculating the actual coordinates of each BGA pad in the BGA pad array, includes: Measure the center distance between adjacent BGA pads in the effective pad outline one by one, compare each center distance with the standard BGA ball pitch, and count the number of successfully matched pad pairs. When the number of successfully matched pad pairs exceeds the preset number, the BGA pad array is confirmed by fitting calculations based on all standard BGA ball pitches. For each BGA pad in the BGA pad array, the center position of each BGA pad is calculated using a grayscale centroid algorithm, and the thickness gradient variation law of the BGA pad at the center position is analyzed to output the actual coordinates of each BGA pad.

[0009] Optionally, in a fourth implementation of the first aspect of the present invention, the step of constructing the coordinate transformation matrix between the actual coordinates and the design coordinates, and creating a serpentine printing path for the second solder paste printing based on the coordinate transformation matrix, includes: Histogram equalization is performed on the pad area corresponding to the actual coordinates to obtain the sub-pixel coordinates of the pad center. The positional deviation is calculated by comparing the subpixel coordinates of the center of the pad with the design coordinates corresponding to the BGA pad in the PCBA design file to obtain the first positional deviation data. Based on the first position deviation data, perform least squares calculation to solve for the coordinate transformation matrix; The target position coordinates for the second solder paste printing are corrected based on the coordinate transformation matrix, and a serpentine printing path is planned based on the target position coordinates.

[0010] Optionally, in a fifth implementation of the first aspect of the present invention, the step of correcting the target position coordinates of the second solder paste printing according to the coordinate transformation matrix and planning a serpentine printing path based on the target position coordinates includes: The coordinate transformation matrix is ​​used to perform coordinate transformation calculations on the design coordinates of the BGA pads in the PCBA design file to obtain the target position coordinates after the second solder paste printing correction. Establish a mapping relationship between the printing head and the target position coordinates; According to the mapping relationship, the BGA pad array is grouped by row and a serpentine printing path is set for left-right scanning of odd rows and right-left scanning of even rows.

[0011] Optionally, in a sixth implementation of the first aspect of the present invention, the step of grouping the BGA pad array by row according to the mapping relationship and setting a serpentine printing path with odd rows scanning left and right and even rows scanning right and left includes: Based on the row and column position information of each BGA pad in the mapping relationship, the BGA pad array is grouped by row to generate grouped data; Based on the grouped data, a forward scanning direction from left to right is set for the odd-numbered row pad sequence, and a reverse scanning direction from right to left is set for the even-numbered row pad sequence, thus establishing a row scanning direction control table; Calculate the continuous trajectory sequence between adjacent BGA pads according to the row scan direction control table; Boundary collision detection is performed on the continuous trajectory sequence to determine whether each trajectory point in the continuous trajectory sequence exceeds the boundary of the BGA area. At the same time, it is checked whether the distance between the continuous trajectory sequence and the pads of ordinary devices meets the safety clearance, and a serpentine printing path is output.

[0012] Optionally, in a seventh implementation of the first aspect of the present invention, the step of detecting the solder paste state parameters on the PCBA surface and generating a start signal when the solder paste state parameters meet the printing conditions includes: The solder paste viscosity, solder paste temperature and solder paste thickness on the PCBA surface are monitored, and solder paste status parameters are collected in real time. A state assessment model is established with solder paste viscosity, solder paste temperature and solder paste thickness as variables, and the printability index corresponding to the solder paste state parameters is calculated using the state assessment model. Check whether the printability index meets the printing conditions. When the printing conditions are met, confirm that the optimal printing state has been reached, and generate a start signal containing a timestamp and a serpentine printing path based on the optimal printing state.

[0013] Optionally, in an eighth implementation of the first aspect of the present invention, receiving the start signal, performing a second solder paste printing according to the serpentine printing path and performing real-time compensation to obtain a printed PCBA includes: The timestamp and serpentine printing path in the start signal are analyzed, the second solder paste printing equipment is started and the stereo vision system is turned on simultaneously, and solder paste is printed on the pad area according to the serpentine printing path; The stereo vision system is used to continuously monitor the relative positional relationship between the printing head and the BGA pads. After printing each BGA pad, the second positional deviation data is calculated based on the relative positional relationship. Based on the second position deviation data, Kalman filtering is performed to predict the position deviation of the next printing point. The position deviation is then input into a PID feedforward compensation controller for real-time compensation, resulting in a printed PCBA.

[0014] This invention also provides a solder paste dual-printing device for PCBA surface mount technology, comprising: The scanning module is used to scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data, and extract the actual coordinates of the BGA pads from it. The construction module is used to construct the coordinate transformation matrix between the actual coordinates and the design coordinates, and to create the serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; The generation module is used to detect the solder paste state parameters on the PCBA surface, and generate a start signal when the solder paste state parameters meet the printing conditions. The second solder paste printing module is used to receive the start signal, perform the second solder paste printing according to the serpentine printing path and perform real-time compensation to obtain the printed PCBA.

[0015] In summary, this invention acquires solder paste surface morphology data using laser 3D scanning technology. Combined with a pad recognition algorithm based on solder paste thickness differences and a ball-to-ball spacing verification mechanism, it can accurately identify the actual distribution position of BGA pads, avoiding the limitations of traditional 2D image recognition methods. By employing sub-pixel-level edge detection and affine transformation matrix establishment, it can accurately describe the geometric relationship between the actual coordinate system and the design coordinate system, effectively compensating for the impact of PCB manufacturing tolerances and component placement deviations on printing accuracy. Through a serpentine scanning mode and mapping relationship establishment, intelligent optimization of the print head's movement trajectory is achieved, reducing invalid movement time. Simultaneously, boundary collision detection ensures that printing actions are strictly limited to the BGA area, avoiding misprinting and contamination of ordinary components. A solder paste surface state evaluation model with viscosity, temperature, and thickness as variables is established. Real-time monitoring of multiple parameters enables precise control of the timing of the second printing, ensuring good fusion of the two layers of solder paste. Using a Kalman filter prediction algorithm and PID feedforward compensation control, it can predict and compensate for positional deviations during the printing process, significantly improving the alignment accuracy and consistency of double printing. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the solder paste double printing method steps in a PCBA surface mount process according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a solder paste dual printing equipment for PCBA surface mount technology in one embodiment of the present invention; Figure 3 This is a schematic diagram of a serpentine printing path in one embodiment of the present invention.

[0017] 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

[0018] 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.

[0019] Reference Figure 1 This embodiment provides a method for dual solder paste printing in PCBA surface mount technology, including the following steps: S11, Scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data and extract the actual coordinates of BGA pads from it; The process involves activating a laser-based height measurement device after the first solder paste printing. This device uses a 635nm red laser and controls the laser scanning head to move at a constant speed of 5mm / s along the X-axis and step along the Y-axis at 10µm intervals to perform a three-dimensional contour scan of the entire PCBA surface, acquiring raw point cloud data in X, Y, and Z dimensions. To address measurement noise during the scanning process, a Gaussian filter with a standard deviation of 0.5 is applied to the three-dimensional coordinate data for preprocessing. A 3×3 pixel filter window is used to smooth the Z-axis height values, resulting in a noise-optimized height measurement dataset. Based on this dataset, exposed substrate areas not covered by solder paste are selected on the PCBA surface. The height of points within these exposed substrate areas is used as the calculation benchmark. A benchmark plane equation, Z=ax+by+c, is established through least-squares fitting. This allows for the relativization of the height of each scanned point, calculating the relative height difference between the actual measured height and the benchmark plane height. The relative heights of all measured points are then reorganized and reconstructed into a topographic image with spatial distribution characteristics. The component layout diagram in the PCBA design file is used to map the coordinates of BGA devices, ordinary surface mount pads, and areas without components defined in the design onto the topographic image. The topographic image is then divided into three categories based on spatial topology: the expected location area for BGA devices, the area for ordinary device pads, and the blank area of ​​the substrate. Based on the solder paste thickness, shape characteristics, and array regularity, the positional contours of the solder paste coverage within the BGA area are identified. Morphological filtering and grayscale centroid calculation are used to extract the center coordinates of each BGA pad, and these coordinates are output as the actual coordinates of the BGA pads after the first printing.

[0020] S12, construct the coordinate transformation matrix between the actual coordinates and the design coordinates, and create the serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; Specifically, image processing enhancement operations are performed on the pad area image. Adaptive histogram equalization is applied to the local image region of each pad to expand the dynamic range of the original grayscale image to 0 to 255, thereby improving the grayscale contrast between the pad edges and the background. A sub-pixel-level edge detection algorithm based on Zernike moments is used to refine each pad area, extracting pad edge information, and the sub-pixel coordinates of the center of each pad are determined through local grayscale centroid calculation. The sub-pixel-level actual center coordinates are then matched point-by-point with the preset theoretical coordinates of the BGA pads in the PCBA design file. By calculating the displacement difference between each pair of coordinates in the X and Y axes, a first position deviation data set containing error components such as global scaling, rotation, and translation is formed. Based on the first position deviation data, a mathematical model for coordinate transformation is constructed using the least squares method. By minimizing the sum of squares of all pad position errors, the six coefficients of the affine transformation matrix are solved, obtaining the coordinate transformation matrix that maps the design coordinates to the actual coordinates. The coordinate transformation matrix includes rotation, scaling, tilting, and translation terms, reflecting the geometric distortions caused by process tolerances, equipment errors, or positioning deviations during the printing process. The target position coordinates used for motion control during the second printing process are input into the coordinate transformation matrix for mapping correction, ensuring that all target point coordinates more closely match the actual solder paste distribution after the first printing. Based on the transformed set of target coordinates and the layout rules of the BGA pad array in the row and column directions, a serpentine path strategy is used for trajectory planning. This involves sequentially scanning each pad row by row, arranging the path from left to right for odd-numbered rows and from right to left for even-numbered rows, to achieve a continuous and smooth motion trajectory design. A linear connection path is planned between each pair of adjacent coordinates, and the movement speed is controlled according to the standard spacing between pads to ensure that the precise alignment of the next point is not affected during high-speed movement, thus forming the second printing trajectory.

[0021] S13 detects the solder paste status parameters on the PCBA surface. When the solder paste status parameters meet the printing conditions, a start signal is generated. It should be noted that the solder paste viscosity, temperature, and thickness on the PCBA surface are continuously monitored, and the collected solder paste status parameters are uploaded to the central control system in real time. Solder paste viscosity is obtained using a non-contact viscosity measurement device based on the laser Doppler velocimetry principle. This device uses a 2MHz ultrasonic probe to irradiate the solder paste surface and measures the frequency shift of the reflected wave, thereby deriving the surface microflow velocity and converting it into the current viscosity value. Solder paste temperature is measured in real time by an infrared thermal imager with a detection band between 8 and 14 micrometers and a temperature resolution accurate to 0.1°C, enabling the creation of temperature field maps. Solder paste thickness monitoring relies on laser displacement sensors, with multiple monitoring points deployed at key BGA pad locations, achieving an accuracy of 0.1 micrometers, thus ensuring precise perception of the solder paste accumulation state at each location. Based on the aforementioned data stream, a multi-factor state evaluation model is constructed with viscosity η(t), temperature T(t), and thickness h(t) as input variables. Through weighted combination, normalization, and time-series analysis of multiple physical state variables, a printability index I(t) comprehensively characterizes the current printability of the solder paste. A preset printability threshold I is used. th This is used to determine whether the current state has entered the optimal printable window, when I(t)≥I th If the solder paste viscosity has not exceeded the set upper limit, the temperature is within ±2°C of the environmental reference value, the thickness has not shown significant collapse, and the surface wettability is acceptable, then the optimal state for entering the second printing is determined. A start signal containing a timestamp and a serpentine printing path is generated based on the optimal printing state.

[0022] S14: Receive the start signal, perform the second solder paste printing according to the serpentine printing path and perform real-time compensation to obtain the printed PCBA.

[0023] Specifically, the system receives and parses the timestamp and serpentine printing path data contained in the start signal. After confirming that the current system clock is consistent with the start time and completing the path parsing of the printing trajectory file, the second solder paste printing equipment is started, and the stereo vision measurement system mounted in front of and to the side of the print head is activated simultaneously. The stereo vision measurement system has a field of view of 20mm×20mm and a resolution of 2μm, and can capture the relative position information of the print head and the pad position in real time during the execution process. Under the serpentine path, the printing equipment prints the solder paste sequentially according to the arrangement order of the pads corrected by affine coordinate transformation. After the solder paste transfer action of each BGA pad is completed, the stereo vision system immediately captures the actual landing point image of the current print head at the coordinate point, and calculates the offset between the current printing point and the design target point in the X and Y axes using the normalized cross-correlation template matching method, forming a second position deviation dataset. Using the second position deviation data as input, continuous time series prediction is performed through the Kalman filter algorithm. By utilizing the evolution trend of the historical deviation vector, the position deviation of the next pad printing point is estimated, and the predicted value is output in the form of a state vector. The predicted deviation is input into the PID feedforward compensation controller. The control quantity is calculated according to the proportional, integral and derivative parameters Kp=0.8, Ki=0.2 and Kd=0.1. The real-time coordinate correction is output and superimposed on the original trajectory coordinates to generate the real-time control coordinates for the execution of the current printing instruction, thus obtaining the printed PCBA.

[0024] In one example, the PCBA surface after the first solder paste printing is scanned to obtain solder paste topography data and extract the actual coordinates of the BGA pads, including: The laser is activated to measure the height of the PCBA surface after the first solder paste printing is completed, and three-dimensional coordinate data is obtained. Gaussian filtering is applied to the three-dimensional coordinate data to generate height measurement data; The exposed area of ​​the PCBA substrate in the height measurement data is selected as the reference plane, and the height difference of each measurement point relative to the reference plane is calculated to generate a topographic image. The layout information in the PCBA design file is matched and marked with the topographic image to identify the expected location area of ​​BGA devices, the pad area of ​​ordinary devices and the blank area of ​​the substrate, and integrated into solder paste topographic data. Extract the actual coordinates of BGA pads from solder paste morphology data.

[0025] In this example, a 3D height scanning system is initiated after the first solder paste printing process. This system integrates a 635nm red laser emitter and a high-speed optical receiver module. Under unified control, the laser scanning head moves at a constant speed of 5mm / s along the X-axis of the PCBA surface and performs line-by-line scanning along the Y-axis at 10µm increments, forming a dense point cloud measurement process covering the entire PCB area. Each scan point is recorded with its X and Y coordinates and corresponding Z-axis height value, forming a 3D spatial coordinate dataset D(x, y, z), reflecting the height distribution of the solder paste surface at the micrometer level. Gaussian filtering is applied to the 3D coordinate data using a 2D Gaussian filter kernel with a standard deviation σ of 0.5 and a window size of 3×3 pixels. Local weighted smoothing is performed on the Z-values ​​in each point cloud, effectively eliminating high-frequency noise and isolated point disturbances caused by measurement errors, resulting in more continuous and trend-clearer height measurement data. Exposed substrate areas, i.e., the base copper foil surface not covered by solder paste and without pads and components, are identified on the PCBA surface. These areas appear as continuous low-lying planes with stable height values ​​in the height map. Based on component layout information or image gradient analysis, these exposed areas are spatially labeled, and their height values ​​are used as input. A reference plane function Z = ax + by + c is constructed using the least squares fitting method, where a, b, and c are the optimal fitting coefficients, and the reference plane is defined as a reference height plane. The relative height is calculated by subtracting the actual Z values ​​of all scan points from the theoretical height values ​​of the corresponding positions on the reference plane, thereby eliminating the overall deformation error caused by board warping or printing angle deviations, and forming a solder paste topography image. The component layout and package size data in the PCBA design file are projected onto the topography image, and the expected location areas of BGA components, ordinary component pad areas, and blank substrate areas without component coverage are marked according to the design coordinate system. After spatial alignment, each topography image point is assigned a region label, forming a solder paste topography dataset containing coordinates, height, time, and category attributes. Based on the solder paste morphology dataset, a pad recognition algorithm for BGA regions was launched. Since the solder paste thickness of BGA pads is higher than the surrounding substrate after printing, all points exceeding a set threshold (e.g., 15 μm) were extracted. Morphological closing and opening operations were used to process the pad contours, removing burrs and filling holes. The Canny edge detection algorithm was used to extract the pad edges. Circular regions with an area between 0.04 and 0.16 square millimeters were identified as valid pad candidates through contour area filtering. The grayscale centroid of each pad candidate region was extracted, and its geometric centroid coordinates were calculated, which were then used as the actual coordinates of the pad.

[0026] In one example, the actual coordinates of the BGA pads are extracted from the solder paste topography data, including: Using a pre-set thickness threshold, high-thickness pixels exceeding the thickness threshold within the expected location area of ​​the BGA device are screened from the solder paste morphology data to form a candidate pad distribution map; The candidate pad distribution map is opened and closed using a circular structuring element to obtain the effective pad outline. Measure the spacing between effective pad outlines and match it with the standard BGA ball pitch to confirm the BGA pad array, and calculate the actual coordinates of each BGA pad in the BGA pad array.

[0027] In this example, local feature extraction is performed on the region marked as the expected location of the BGA device based on the 3D topography data of solder paste. A thickness threshold is set for the BGA region to distinguish the solder paste-covered area from the background substrate area. The thickness threshold is determined based on the typical BGA pad printing characteristics. All topography data points within the BGA region are traversed. When the relative height value of a point is greater than the thickness threshold, it is determined to belong to the BGA pad and marked as a candidate region pixel. All high-thickness pixels that meet the condition are collected to form a preliminary candidate pad distribution map. Morphological filtering is performed on the candidate pad distribution map, using a circular structuring element with a diameter of 20μm, and opening and closing operations are performed sequentially. The opening operation removes noise, broken edges, and isolated points, while the closing operation fills the voids inside the pad region, making the pad outline more rounded and the edges continuous. After filtering, effective pad outlines with clear boundaries and stable areas are obtained. Each effective pad outline is geometrically represented as a high-thickness region that is approximately circular or elliptical. Its boundary is extracted and vectorized using an image contour extraction algorithm (such as the Canny edge detector). To verify whether these contours truly correspond to the pads in the BGA array, an array consistency matching mechanism is introduced. Using the package parameters of the BGA device as a standard, a list of common standard ball pitches is preset, such as 0.4mm, 0.5mm, 0.65mm, 0.8mm, and 1.0mm. The Euclidean distance is calculated for the centroid coordinates of each pair of adjacent pads to obtain the actual ball pitch. By matching the actual pitch with the standard value and setting a matching tolerance (e.g., ±10μm), pad pairs with stable arrangement characteristics are selected to construct a multi-row, multi-column two-dimensional array structure. The regularity of the overall arrangement is evaluated by calculating the slope of the fitted line for each row and column and checking whether the angular deviation is controlled within ±2°. The set of pads that meets the above geometric consistency conditions is confirmed as the BGA pad array. After confirming the array, the center point position is re-extracted within each valid pad area. A grayscale centroid algorithm is used to weight all pixels within the area according to their corresponding height to calculate the sub-pixel level geometric center of the pad. The actual coordinates of all BGA pads are stored uniformly.

[0028] The process, after generating the candidate pad distribution map, includes BGA pad array feature recognition and dual verification based on solder paste surface height differences. This includes: performing morphological gradient processing on each candidate pad region in the candidate pad distribution map; using a 20μm diameter circular structuring element to perform opening and closing operations to eliminate solder paste surface burrs and fill holes; subsequently applying the Canny edge detection operator to extract pad edge contours; and filtering effective contours by area range to obtain regular pad contour feature data. Based on the pad contour feature data, a BGA pad ball pitch standard matrix verification model is established, and the identified pad centroid coordinates are compared with five standard BGA ball pitches: 0.4mm, 0.5mm, 0.65mm, 0.8mm, and 1.0mm. Row matching calculation is performed. When the deviation between the actual spacing and the standard spacing is less than 10μm, it is confirmed as an effective ball spacing, and the ball spacing verification result is obtained. For each effective pad in the ball spacing verification result, the solder paste thickness gradient is double-verified. The solder paste thickness gradient value is calculated along the radial direction of the pad to verify the condition that the thickness gradient in the center area of ​​the pad is less than 2μm / pixel and the thickness gradient in the edge area is greater than 5μm / pixel. At the same time, the angular deviation of the row and column directions of the pad array is checked to be less than 2°, and the thickness gradient verification data is obtained. The thickness gradient verification data and the ball spacing verification result are fused and analyzed. When both the ball spacing matching and thickness gradient distribution conditions are met, it is confirmed as a standard BGA pad array, and the BGA pad array identification result after double verification is output.

[0029] In one example, the spacing between valid pad outlines is measured and matched against standard BGA ball pitch to confirm the BGA pad array, and the actual coordinates of each BGA pad in the BGA pad array are calculated, including: Measure the center distance between adjacent BGA pads in the effective pad outline one by one, compare each center distance with the standard BGA ball pitch, and count the number of successfully matched pad pairs; When the number of successfully matched pad pairs exceeds the preset number, the BGA pad array is confirmed by fitting calculations based on all standard BGA ball pitches. For each BGA pad in the BGA pad array, the grayscale centroid algorithm is used to calculate the center position of each BGA pad, and the thickness gradient variation law of the BGA pad at the center position is analyzed to output the actual coordinates of each BGA pad.

[0030] In this example, the initial center coordinates of each pad region are extracted based on the effective pad outline map. The pad center points are paired one-to-one according to their spatial coordinate distribution, and the Euclidean distance between each pair of pad center points is calculated to obtain a distance matrix representing the distance between adjacent pads. A list of standard BGA package ball pitches is introduced, including typical values ​​such as 0.4 mm, 0.5 mm, 0.65 mm, 0.8 mm, and 1.0 mm, with a tolerance range set (e.g., ±10 micrometers). Each center distance is compared one by one to determine if it falls within a certain standard ball pitch and its tolerance range. If the condition is met, it is considered a valid matching pad pair. The number of successful matches under each ball pitch standard is gradually accumulated, and the total number of matches is counted. When the number of successful matches exceeds a preset lower limit (e.g., the number of valid pairs in the array reaches more than 70% of the total possible pairs), the current pad region is considered to have clear array characteristics and possess spatial consistency of the standard BGA package layout. Based on this, the spacing values ​​of all successfully matched pad pairs are fitted and calculated. The least squares method is used to globally average all center distances to determine the actual ball spacing. A linear fit is then performed on the arrangement direction of the pad center points to obtain the row and column direction vectors of the array, forming a BGA pad array structure model including array direction, array spacing, and array starting point. For each pad region within the BGA pad array, a grayscale centroid algorithm is used to calculate its center point position. The grayscale centroid algorithm is based on the relative height values ​​of the pad regions in the solder paste 3D topography image. It multiplies the spatial position of each pixel with its corresponding height value, sums the results, and then divides by the sum of the total height values ​​of the region to calculate the sub-pixel-level precise coordinates of the pad centroid. To verify the printing consistency and configuration correctness of the pad thickness, the thickness distribution gradient at the center position of each pad is analyzed. Thickness distribution values ​​are extracted within a certain radius (e.g., 20 μm) centered on the center point, and the gradient vector field of the region is calculated. The rate of thickness change at the center is obtained by differentiation. If the thickness gradient at the center is less than 2 μm / pixel, while the thickness gradient at the outer edge is greater than 5 μm / pixel, it indicates that the thickness profile of the pad conforms to the geometry of a typical BGA solder ball, exhibiting a structural feature of a convex center and steeply sloping edges. The grayscale centroid coordinates of each pad are combined with the thickness gradient change results for consistency verification, and the actual location set of the BGA pads is output in coordinate form.

[0031] In this embodiment, after outputting the actual coordinates of each BGA pad, the method further includes reconstructing the actual coordinates of the BGA region based on sub-pixel level detection using machine vision. This includes: performing adaptive histogram equalization enhancement processing on the pad region corresponding to the actual coordinates of each BGA pad to expand the grayscale dynamic range of the pad region to the full range, resulting in a contrast-enhanced pad image; and using a Zernike moment sub-pixel level edge detection algorithm based on the contrast-enhanced pad image to accurately locate the edge contour of each BGA pad. By calculating the 0th to 4th order Zernike moments within a 3×3 pixel window and extracting phase information, the detection accuracy of the pad edge position is improved from the pixel level to the sub-pixel level, resulting in sub-pixel precision pad edge coordinate data. The process involves geometrically matching subpixel precision pad edge coordinate data with the corresponding theoretical BGA pad positions in the PCBA design file. The least squares method is used to calculate the affine transformation relationship between the actual and theoretical coordinate systems, establishing a six-parameter affine transformation matrix that includes scaling, rotation, translation, and tilt transformation parameters. This yields a transformation matrix describing the coordinate system deviation. Based on this transformation matrix, the outer boundary of the BGA pad array is dynamically reconstructed. The convex hull algorithm is used to calculate the convex hull boundary of all pad centroids and expands it outwards by a preset safety margin. Simultaneously, collision detection and spacing verification with ordinary component pads are performed. When the boundary distance from an ordinary component pad is less than the safety spacing, the boundary position is automatically adjusted, ultimately generating accurate dynamic boundary coordinates for the BGA region.

[0032] In one example, a coordinate transformation matrix is ​​constructed between the actual coordinates and the design coordinates, and a serpentine printing path for the second solder paste printing is created based on the coordinate transformation matrix, including: Histogram equalization is performed on the pad area corresponding to the actual coordinates to obtain the sub-pixel coordinates of the pad center; The positional deviation is calculated by comparing the subpixel coordinates of the pad center with the design coordinates corresponding to the BGA pad in the PCBA design file to obtain the first positional deviation data. Based on the first positional deviation data, perform least squares calculation to solve for the coordinate transformation matrix; The target position coordinates for the second solder paste printing are corrected based on the coordinate transformation matrix, and a serpentine printing path is planned based on the target position coordinates.

[0033] In this example, image processing enhancement is performed on the pad areas extracted after the first solder paste printing. For each pad area, a local image window containing the pad area is selected, and histogram equalization is performed on the local image window to expand the grayscale dynamic range to 0 to 255, thereby improving the grayscale contrast between the center and surrounding areas of the pad. A subpixel edge extraction method based on Zernike moments or image phase features is called to scan the pad contour edges pixel by pixel, and the offset of the edge pixels at the subpixel level is calculated in the local area of ​​each pad boundary to locate the coordinates of the pad center. The subpixel-level center coordinates are paired one by one with the preset BGA pad design coordinates in the PCBA design file, and the distance difference between the actual coordinates and the design coordinates in the horizontal and vertical directions is calculated to form the first position deviation dataset. In the first position deviation dataset, each pad point has two offset components, corresponding to the deviations in the X and Y directions, respectively. To establish a global mapping from theoretical to actual coordinates, after obtaining all pad deviation data, an affine model was calculated using the least squares method. An optimal set of parameters was obtained through the error minimization criterion to characterize the overall board deformation, pad offset, rotation, scaling, and other geometric transformations, resulting in the coordinate transformation matrix. The affine model has six parameters, corresponding to scaling, rotation, translation, and tilt components, effectively performing global geometric corrections to the overall coordinate system. Based on the coordinate transformation matrix, the target position coordinates for the second solder paste printing were corrected. All original target position coordinates were derived from the BGA pad standard array in the design file. After unified correction using the coordinate transformation matrix, a set of calibrated coordinates that better fit the actual pad distribution of the first printing was obtained. The calibrated target coordinates constitute the base coordinate set for the second printing. A serpentine path generation program was initiated based on the target position coordinates, sorting the pad array in row-major order. Odd-numbered rows connected points from left to right, and even-numbered rows connected points from right to left, generating a continuous, serpentine path. The serpentine path maintains a constant printing dwell time of 0.3 seconds at the center of each pad. The speed gradually decreases from 80 mm / s in the non-printing section to 20 mm / s within the pad area, coupled with a printing pressure of 2.5 Newtons and a stencil gap of 100 micrometers to ensure stable printing quality and consistent shape. Throughout the path generation process, the path is checked in real-time to ensure it remains within dynamic boundary limits, and path corrections are performed when it approaches the boundaries to avoid misprinting due to the path exceeding the limits or interfering with adjacent pads. The output path file contains the coordinate data of each point, speed parameters, printing pressure value, and timestamp, and is stored in standard G-code format.

[0034] like Figure 3 As shown, Figure 3 This is a schematic diagram of the serpentine printing path in an embodiment of the present invention.

[0035] In one example, the target position coordinates for the second solder paste printing are corrected according to the coordinate transformation matrix, and a serpentine printing path is planned based on the target position coordinates, including: By using a coordinate transformation matrix, the design coordinates of the BGA pads in the PCBA design file are transformed to obtain the target position coordinates after the second solder paste printing correction. Establish a mapping relationship between the print head and the target position coordinates; Based on the mapping relationship, the BGA pad array is grouped by row and a serpentine printing path is set for left-right scanning of odd rows and right-left scanning of even rows.

[0036] In this example, based on the coordinate transformation matrix, a batch transformation operation is performed on the standard coordinates of all BGA pads contained in the PCBA design file to obtain a set of target position coordinates. These target position coordinates are geometrically aligned with the actual landing point of the first solder paste printing and meet the spatial constraints and correction requirements for the second printing. A mapping relationship is established between the printing head working coordinate system and the target position coordinates. This mapping relationship includes the X and Y axis positions in two-dimensional space, and considers the stencil gap height control and the time synchronization mechanism of the printing head's advance and retreat movements along the Z-axis. Within the BGA area boundary, based on the position of each coordinate point, a composite control parameter is generated, containing its X and Y plane coordinates, corresponding printing pressure, expected printing dwell time, and up / down movement commands. This composite control parameter is then bound to each target point coordinate, forming a point-to-point motion control mapping set. Through this mapping mechanism, the printing control system converts each motion path segment into a motor drive signal sequence during actual execution, ensuring precise timing, continuous movement, and accurate positioning between each stage of the printing head's positioning, landing, contact, printing, and lifting processes. To optimize the motion path and improve printing efficiency, all corrected target coordinate points are grouped according to their Y-axis values. Each group of pads with similar Y values ​​is considered a member of the same row, sorted in ascending order along the X-axis, and then assigned row numbers. For path planning, a serpentine path rule is adopted: odd-numbered rows are scanned sequentially from the leftmost coordinate to the right, while even-numbered rows are scanned sequentially from the rightmost coordinate to the left. This eliminates the need for large-scale reverse movements during inter-row transitions, reducing the distance and time spent by the print head during idle movement and improving overall printing cycle efficiency. In the path segment between every two pads, the non-printing segment is uniformly set to a speed of 80 mm / s, automatically slowing to 20 mm / s before entering the effective printing area of ​​each pad, maintaining a 0.3-second printing dwell time at the center point, and applying a printing pressure of 2.5 Newtons to ensure uniform solder paste deposition thickness and stable adhesion. Throughout the path generation process, dynamic boundary detection and minimum safe distance checks are performed on each path segment to ensure that the path does not exceed the BGA area boundary or interfere with adjacent ordinary pad areas. When the path is detected to be about to enter the boundary critical line or the safe area of ​​ordinary solder pads, the trajectory adjustment mechanism is activated to locally reconstruct the path segment, prioritizing positional accuracy and printing safety. The complete path set, including serpentine motion logic, coordinate correction results, control parameters, and timestamps, is exported as standard G-code format for real-time reading and execution by the second solder paste printing equipment.

[0037] As shown in Table 1, Table 1 is a BGA pad coordinate transformation table:

[0038] In one example, the BGA pad array is grouped by row according to the mapping relationship, and a serpentine printing path is set for odd rows to scan left and right and even rows to scan right and left, including: Based on the row and column position information of each BGA pad in the mapping relationship, the BGA pad array is grouped by row to generate grouped data; Based on the grouped data, a forward scan direction from left to right is set for the odd-numbered row pad sequence, and a reverse scan direction from right to left is set for the even-numbered row pad sequence, thus establishing a row scan direction control table; Calculate the continuous trajectory sequence between adjacent BGA pads according to the row scan direction control table; Perform boundary collision detection on the continuous trajectory sequence to determine whether each trajectory point in the continuous trajectory sequence exceeds the boundary of the BGA area. At the same time, check whether the distance between the continuous trajectory sequence and the pads of ordinary components meets the safety clearance and output a serpentine printing path.

[0039] In this example, based on the target location coordinates of the BGA pads, the center point of each pad is extracted in a two-dimensional coordinate system, and all pads are classified into rows according to their height value along the Y-axis. A row grouping threshold is set, for example, the allowable range for the vertical spacing of each row of pads is set to ±20 micrometers. Pads with similar adjacent Y values ​​are searched row by row from top to bottom, and row numbers are assigned sequentially. A data structure with the pad center point as the unit and a row number label is established, forming a grouped data table that reflects the spatial arrangement of each pad in the row direction throughout the BGA array. Different scanning direction strategies are defined for the pad sequence in each row based on the parity of the row number. For all odd-numbered pad rows, the scanning order is set to start from the smallest X-coordinate and move sequentially towards the largest X-coordinate, i.e., a forward scan from left to right. For even-numbered pad rows, the scanning order is set to start from the largest X-coordinate and move sequentially towards the smallest X-coordinate, i.e., a reverse scan from right to left. This creates a "serpentine" scanning order, forming continuous and shortest paths between pads, avoiding large-scale backtracking of the print head after each row, thus improving printing efficiency and reducing the number of trajectory switching. The connection order of the scanning start point, end point, and intermediate pad points of each row is organized by row number and uniformly written into the row scanning direction control table. According to the row scanning direction control table, the coordinate points in the pad sequence of each row are read row by row, and the straight-line connection path between adjacent points is calculated pair by pair. The start point, target point, and connection path points are uniformly generated into trajectory segments, and all trajectory segments are connected to form a continuous trajectory sequence, including coordinate path point information, and accompanying parameters such as the moving speed, printing pressure, and time of each point. To ensure the safety and accuracy of the entire trajectory execution, boundary collision detection is performed on the continuous trajectory sequence. The dynamic boundary of the BGA region, determined during the pad identification stage, is loaded. This dynamic boundary extends 150 micrometers outward from the pad bulge to form a closed polygonal region. During trajectory generation, a point inclusion check is performed on each trajectory point to determine whether it is inside the BGA boundary polygon. If a trajectory point falls outside the boundary, it is marked as a collision point, and a trajectory correction mechanism is initiated. The collision point is projected onto the nearest boundary line segment to form an alternative point for path reconstruction, ensuring that all trajectory points are confined within the effective range of the BGA region. Simultaneously, the minimum safe distance between the trajectory points and ordinary device pads is checked to ensure it meets process requirements.A spatial database of common component pads is pre-established, recording the center coordinates and physical radii of all non-BGA pads. During the trajectory verification phase, the Euclidean distance between each trajectory point and the common pads is calculated. When the Euclidean distance is detected to be less than the safety limit of 200 micrometers, a safety distance violation judgment is triggered, and local path reconstruction is performed. Using an artificial potential field method, common pads are treated as repulsive sources, and BGA pad targets are treated as attractive sources. The movement direction and position of the trajectory points are dynamically adjusted to ensure that the new trajectory points both meet the target travel direction and avoid all obstacle pads. Through the continuous processing of row grouping, scan control, trajectory connection, boundary detection, and safety distance verification, a serpentine printing path is output and recorded in G-code format, containing precise positioning instructions for each pad printing action.

[0040] In one example, the solder paste state parameters on the PCBA surface are detected. When the solder paste state parameters meet the printing conditions, a start signal is generated, including: The solder paste viscosity, solder paste temperature and solder paste thickness on the PCBA surface are monitored, and solder paste status parameters are collected in real time. A state assessment model with solder paste viscosity, solder paste temperature and solder paste thickness as variables was established, and the printability index corresponding to the state parameters of the solder paste was calculated using the state assessment model. Check if the printability index meets the printing conditions. When the printing conditions are met, confirm that the optimal printing state has been reached, and generate a start signal containing a timestamp and a serpentine printing path based on the optimal printing state.

[0041] In this example, after the first solder paste printing is completed, a multi-dimensional physical parameter monitoring system covering the BGA area is activated to collect real-time data on three key parameters of the solder paste surface: viscosity, temperature, and thickness. For viscosity monitoring, a non-contact viscometer based on laser Doppler velocimetry technology is used. Its principle is to excite the solder paste surface with high-frequency ultrasound, specifically in the 2MHz band. By measuring the frequency shift between the reflected and incident waves, the microscopic flow velocity of the solder paste surface is calculated and then converted into a viscosity value. The sensor is set to automatically sample every 10 seconds and transmit the data directly back to the central control system, reflecting the viscosity changes caused by air contact or moisture evaporation. Simultaneously, an infrared thermal imager array is deployed to perform a full-view scan of the PCB surface temperature distribution. The infrared thermal imager has a detection wavelength range of 8 to 14 micrometers and a resolution of 0.1°C, ensuring timely identification of temperature anomalies as the solder paste gradually evaporates or heat dissipates. In critical pad areas, multiple laser displacement sensor nodes are deployed. These sensors, with a resolution of 0.1 micrometers, record the thickness change trend of the solder paste during free expansion. This is suitable for monitoring localized collapse and thickness loss caused by gravity, surface tension, and wetting properties in the short period after printing. As the monitoring data is continuously updated, a built-in state assessment model is invoked, using solder paste viscosity η, temperature T, and thickness h as three core input variables to calculate the printability index I at the current moment. The state assessment model integrates the empirical weighting relationships between multiple influencing factors and material dynamics characteristics, quantifying the deviation of the current printability state of the solder paste from the standard process window. Using SAC305 lead-free solder paste as a benchmark, when η is close to 150 Pa·s, T is stable within ±2°C of ambient temperature, and h remains above 90% of the original thickness, the printability index I approaches 1, indicating that it is in a completely ideal printable state. Conversely, if any variable exceeds the critical fluctuation range, the I value will drop rapidly, indicating that it is not suitable to immediately perform a second printing operation. After each model update, a conditional check is performed on the I value. If the I value is greater than a preset threshold I... th If the state remains unchanged for more than a specified time (e.g., 30 seconds) without any fluctuation trend, this state is defined as reaching the optimal printing window, and the printing process preparation logic is triggered. The serpentine printing path file previously generated based on the actual BGA coordinates is read, and the path file is bound to the current system timestamp to generate a printing start signal. This signal contains the start time information under the current optimal state, and embeds specific path structure data, target coordinate sequence, and speed / pressure control parameters.

[0042] The process includes real-time acquisition of solder paste state parameters, followed by intelligent judgment of the solder paste printability time window based on multi-parameter fusion. This involves: using laser Doppler velocimetry to emit 2MHz ultrasonic waves onto the solder paste surface and detecting the frequency shift of the reflected waves; combining infrared thermal imager temperature monitoring and laser displacement sensor thickness measurement to establish a real-time evaluation model of the solder paste surface state; and calculating the comprehensive solder paste printability index by fusing viscosity, temperature, and thickness parameters. Based on the comprehensive solder paste printability index, a high-speed camera is used to observe the solder paste surface wetting behavior, calculate the contact angle change rate, and analyze the surface tension change trend. Simultaneously, X-ray photoelectron spectroscopy is used to detect the solder paste surface wetting behavior. The thickness of the oxide layer on the solder paste surface is used to obtain auxiliary evaluation parameters for the physicochemical state of the solder paste surface. The comprehensive index of solder paste printability and the auxiliary evaluation parameters are input into the time window judgment algorithm for multi-condition verification. When the viscosity value is lower than the preset threshold, the temperature is maintained within ±2°C of the ambient temperature, the contact angle change rate is greater than -0.5° / s, and the oxide layer thickness is less than 2nm, the solder paste is confirmed to have reached the optimal printable time window. Based on the confirmation result of the optimal printable time window, the start signal generation program is immediately triggered. The current timestamp, solder paste state parameters, and serpentine printing path information are encapsulated into a start signal data packet, and a precise execution instruction is sent to the second solder paste printing system.

[0043] In one example, upon receiving a start signal, a second solder paste printing is performed following a serpentine printing path with real-time compensation, resulting in a completed PCBA, including: Analyze the timestamp and serpentine printing path in the start signal, start the second solder paste printing equipment and simultaneously turn on the stereo vision system, and print solder paste on the pad area according to the serpentine printing path; A stereo vision system is used to continuously monitor the relative positional relationship between the print head and the BGA pads. After printing each BGA pad, the second positional deviation data is calculated based on the relative positional relationship. Based on the second position deviation data, Kalman filtering is used to predict the position deviation of the next printing point. The position deviation is then input into the PID feedforward compensation controller for real-time compensation, resulting in the printed PCBA.

[0044] In this example, after detecting that the printability index has reached the start condition, the timestamp information and the corresponding serpentine printing path file path are extracted from the start signal. The time control module compares the current system time with the timestamp field of the start signal. After confirming that the synchronization reaches the millisecond level, the path file is called and loaded with the target coordinate sequence, point scanning order, printing pressure, speed control parameters, and runtime parameters contained therein, completing the initialization of the action preparation process. After the path file is loaded, a command is sent to start the second solder paste printing device, and the dual-camera stereo vision measurement system located in front of and to the side of the print head is simultaneously activated. The dual-camera stereo vision measurement system has a 20 mm × 20 mm field of view and a 2 micrometer spatial resolution, enabling real-time acquisition of the relative position images between the print head and the pads during the print head's movement. After printing begins, the device prints sequentially from the leftmost side of the odd-numbered rows to the right, following the serpentine trajectory in the path. After entering the even-numbered rows, it moves from the rightmost side to the left, gradually completing the printing action of the entire array of BGA pads according to the top-down row sequence. After the print head completes the printing action for each pad according to the path, the stereo vision system acquires the actual printing landing point image in real time and compares it with the target coordinates of the pad in the original path. A normalized cross-correlation image matching algorithm is used to identify the spatial offset between the actual landing point of the print head and the target point in the X and Y axes, thereby calculating the second position deviation data, which is the difference between the actual printing position of the current pad and the preset trajectory. After continuously recording the second position deviation data of multiple pad points, a dynamic prediction mechanism based on Kalman filtering is activated. The current deviation data is used as the state observation input, and the state is recursively extrapolated by combining the deviation change trend of the previous printing point. A state vector modeling method including the deviation position and the deviation change rate is used to estimate the deviation evolution trend and output the deviation prediction value for the target point of the next printing action. The position deviation value predicted by Kalman filtering is input into the PID feedforward compensation controller, which activates the three-segment parameter calculation module to perform real-time control quantity calculations on the proportional, integral, and derivative parts, respectively. The proportional coefficient is set to 0.8, the integral coefficient to 0.2, and the derivative coefficient to 0.1, forming the trajectory correction value. The trajectory correction value is directly added to the target coordinate position in the current control command, thus compensating for and adjusting the position of the next target point before the printing action is executed. The entire compensation control loop will be executed immediately after each pad is printed, forming a continuous mechanism of closed-loop feedback and predictive correction. This effectively eliminates trajectory drift caused by motion inertia, equipment flexibility, and cumulative errors in path execution, resulting in a printed PCBA. The process, following the calculation of the second position deviation data, includes dynamic compensation of dual printing trajectories based on Kalman filter prediction. This involves: constructing a four-dimensional state vector from the second position deviation data, including X-direction deviation, Y-direction deviation, the rate of change of X-direction deviation, and the rate of change of Y-direction deviation; establishing a Kalman filter state transition model; describing the temporal variation of deviation through state equations and observation equations to obtain a deviation prediction model; using a recursive calculation method based on the deviation prediction model to predict and analyze the position deviation of the next BGA pad printing point; combining the statistical characteristics of historical deviation data and the current deviation change trend to calculate the confidence interval and error covariance of the predicted deviation, obtaining a high-precision position deviation prediction value; inputting the position deviation prediction value into a PID feedforward compensation controller for trajectory correction calculation; using a proportional, integral, and derivative control strategy to calculate the trajectory correction amount in the X and Y axes; simultaneously combining the mechanical characteristics of the printing head for inertial compensation and vibration suppression to obtain a comprehensive trajectory compensation command; adjusting the position parameters and motion trajectory of the printing head in real time according to the comprehensive trajectory compensation command; and controlling the printing positioning error of each BGA pad within ±3μm through closed-loop feedback control to complete high-precision dynamic compensation of dual printing trajectories.

[0045] Reference Figure 2 This embodiment provides a solder paste dual printing equipment for PCBA surface mount technology, including: Scanning module 21 is used to scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data and extract the actual coordinates of BGA pads from it; Module 22 is used to construct the coordinate transformation matrix between the actual coordinates and the design coordinates, and to create the serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; The generation module 23 is used to detect the solder paste status parameters on the PCBA surface. When the solder paste status parameters meet the printing conditions, a start signal is generated. The second solder paste printing module 24 is used to receive the start signal, perform the second solder paste printing according to the serpentine printing path and perform real-time compensation to obtain the printed PCBA.

[0046] In this embodiment, the specific implementation of each unit in the above device embodiment is described in the above method embodiment, and will not be repeated here.

[0047] This invention acquires solder paste surface morphology data using laser 3D scanning technology. Combined with a pad recognition algorithm based on solder paste thickness differences and a ball-to-ball spacing verification mechanism, it can accurately identify the actual distribution position of BGA pads, avoiding the limitations of traditional 2D image recognition methods. Subpixel-level edge detection and affine transformation matrix establishment accurately describe the geometric relationship between the actual coordinate system and the design coordinate system, effectively compensating for the impact of PCB manufacturing tolerances and component placement deviations on printing accuracy. Through a serpentine scanning mode and mapping relationship establishment, intelligent optimization of the print head's movement trajectory is achieved, reducing invalid movement time. Simultaneously, boundary collision detection ensures that printing actions are strictly limited to the BGA area, avoiding misprinting and contamination of ordinary components. A solder paste surface state evaluation model with viscosity, temperature, and thickness as variables is established. Real-time monitoring of multiple parameters enables precise control of the timing of the second printing, ensuring good fusion of the two layers of solder paste. Using a Kalman filter prediction algorithm and PID feedforward compensation control, it can predict and compensate for positional deviations during the printing process, significantly improving the alignment accuracy and consistency of double printing.

[0048] 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.

[0049] 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 dual-printing of solder paste in PCBA surface mount technology, characterized in that, include: Scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data, and extract the actual coordinates of the BGA pads from it; Construct a coordinate transformation matrix between the actual coordinates and the design coordinates, and create a serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; The solder paste state parameters on the PCBA surface are detected, and a start signal is generated when the solder paste state parameters meet the printing conditions. Upon receiving the start signal, the second solder paste printing is performed according to the serpentine printing path, and real-time compensation is performed to obtain the printed PCBA.

2. The solder paste dual printing method for PCBA surface mount technology according to claim 1, characterized in that, The process of scanning the PCBA surface after the first solder paste printing is completed, obtaining solder paste morphology data, and extracting the actual coordinates of the BGA pads includes: The laser is activated to measure the height of the PCBA surface after the first solder paste printing is completed, and three-dimensional coordinate data is obtained. The three-dimensional coordinate data is subjected to Gaussian filtering to form height measurement data; The exposed area of ​​the PCBA substrate in the height measurement data is selected as the reference plane, and the height difference of each measurement point relative to the reference plane is calculated to generate a topographic image. The layout information in the PCBA design file is matched and marked with the topographic image to identify the expected location area of ​​BGA devices, the pad area of ​​ordinary devices and the blank area of ​​the substrate, and integrated into solder paste topographic data. The actual coordinates of the BGA pads are extracted from the solder paste morphology data.

3. The solder paste dual printing method for PCBA surface mount technology according to claim 2, characterized in that, Extracting the actual coordinates of the BGA pads from the solder paste morphology data includes: Using a pre-set thickness threshold, high-thickness pixels exceeding the thickness threshold are selected from the solder paste morphology data within the expected location area of ​​the BGA device to form a candidate pad distribution map; The candidate pad distribution map is opened and closed using circular structuring elements to obtain the effective pad outline. The spacing between the effective pad outlines is measured and matched with the standard BGA ball spacing to confirm the BGA pad array, and the actual coordinates of each BGA pad in the BGA pad array are calculated.

4. The solder paste dual printing method for PCBA surface mount technology according to claim 3, characterized in that, The process of measuring the spacing between the effective pad outlines and matching it with the standard BGA ball pitch to confirm the BGA pad array, and calculating the actual coordinates of each BGA pad in the BGA pad array, includes: Measure the center distance between adjacent BGA pads in the effective pad outline one by one, compare each center distance with the standard BGA ball pitch, and count the number of successfully matched pad pairs. When the number of successfully matched pad pairs exceeds the preset number, the BGA pad array is confirmed by fitting calculations based on all standard BGA ball pitches. For each BGA pad in the BGA pad array, the center position of each BGA pad is calculated using a grayscale centroid algorithm, and the thickness gradient variation law of the BGA pad at the center position is analyzed to output the actual coordinates of each BGA pad.

5. The solder paste dual printing method for PCBA surface mount technology according to claim 1, characterized in that, The process of constructing the coordinate transformation matrix between the actual coordinates and the design coordinates, and creating the serpentine printing path for the second solder paste printing based on the coordinate transformation matrix, includes: Histogram equalization is performed on the pad area corresponding to the actual coordinates to obtain the sub-pixel coordinates of the pad center. The positional deviation is calculated by comparing the subpixel coordinates of the center of the pad with the design coordinates corresponding to the BGA pad in the PCBA design file to obtain the first positional deviation data. Based on the first position deviation data, perform least squares calculation to solve for the coordinate transformation matrix; The target position coordinates for the second solder paste printing are corrected based on the coordinate transformation matrix, and a serpentine printing path is planned based on the target position coordinates.

6. The solder paste dual printing method for PCBA surface mount technology according to claim 5, characterized in that, The step of correcting the target position coordinates for the second solder paste printing according to the coordinate transformation matrix, and planning a serpentine printing path based on the target position coordinates, includes: The coordinate transformation matrix is ​​used to perform coordinate transformation calculations on the design coordinates of the BGA pads in the PCBA design file to obtain the target position coordinates after the second solder paste printing correction. Establish a mapping relationship between the printing head and the target position coordinates; According to the mapping relationship, the BGA pad array is grouped by row and a serpentine printing path is set for left-right scanning of odd rows and right-left scanning of even rows.

7. The solder paste dual printing method for PCBA surface mount technology according to claim 6, characterized in that, The step of grouping the BGA pad array by row according to the mapping relationship and setting a serpentine printing path with odd rows scanning left and right and even rows scanning right and left includes: Based on the row and column position information of each BGA pad in the mapping relationship, the BGA pad array is grouped by row to generate grouped data; Based on the grouped data, a forward scanning direction from left to right is set for the odd-numbered row pad sequence, and a reverse scanning direction from right to left is set for the even-numbered row pad sequence, thus establishing a row scanning direction control table; Calculate the continuous trajectory sequence between adjacent BGA pads according to the row scan direction control table; Boundary collision detection is performed on the continuous trajectory sequence to determine whether each trajectory point in the continuous trajectory sequence exceeds the boundary of the BGA area. At the same time, it is checked whether the distance between the continuous trajectory sequence and the pads of ordinary devices meets the safety clearance, and a serpentine printing path is output.

8. The solder paste dual printing method for PCBA surface mount technology according to claim 1, characterized in that, The detection of solder paste state parameters on the PCBA surface, and the generation of a start signal when the solder paste state parameters meet the printing conditions, includes: The solder paste viscosity, solder paste temperature and solder paste thickness on the PCBA surface are monitored, and solder paste status parameters are collected in real time. A state assessment model is established with solder paste viscosity, solder paste temperature and solder paste thickness as variables, and the printability index corresponding to the solder paste state parameters is calculated using the state assessment model. Check whether the printability index meets the printing conditions. When the printing conditions are met, confirm that the optimal printing state has been reached, and generate a start signal containing a timestamp and a serpentine printing path based on the optimal printing state.

9. The solder paste dual printing method for PCBA surface mount technology according to claim 1, characterized in that, The process of receiving the start signal, performing a second solder paste printing according to the serpentine printing path and performing real-time compensation to obtain a printed PCBA includes: The timestamp and serpentine printing path in the start signal are analyzed, the second solder paste printing equipment is started and the stereo vision system is turned on simultaneously, and solder paste is printed on the pad area according to the serpentine printing path; The stereo vision system is used to continuously monitor the relative positional relationship between the printing head and the BGA pads. After printing each BGA pad, the second positional deviation data is calculated based on the relative positional relationship. Based on the second position deviation data, Kalman filtering is performed to predict the position deviation of the next printing point. The position deviation is then input into a PID feedforward compensation controller for real-time compensation, resulting in a printed PCBA.

10. A solder paste dual-printing equipment for PCBA surface mount technology, characterized in that, The step of the method for implementing the solder paste dual printing method for the PCBA surface mount process according to any one of claims 1 to 9, wherein the solder paste dual printing equipment for the PCBA surface mount process comprises: The scanning module is used to scan the PCBA surface after the first solder paste printing is completed, obtain solder paste morphology data, and extract the actual coordinates of the BGA pads from it. The construction module is used to construct the coordinate transformation matrix between the actual coordinates and the design coordinates, and to create the serpentine printing path for the second solder paste printing based on the coordinate transformation matrix; The generation module is used to detect the solder paste state parameters on the PCBA surface, and generate a start signal when the solder paste state parameters meet the printing conditions. The second solder paste printing module is used to receive the start signal, perform the second solder paste printing according to the serpentine printing path and perform real-time compensation to obtain the printed PCBA.

Images