A method for improving PCB board warping

Abstract

The present application relates to the technical field of PCB intelligent detection, in particular to a PCB warping improvement method. The method comprises: scanning the PCB placed on the quality inspection platform through a laser scanner to obtain a spectrum pattern; analyzing the deformation index of the PCB based on the spectrum pattern, and adjusting the production process of the PCB based on the deformation index when the deformation index is greater than a preset threshold. The present application obtains the spectrum pattern through the laser scanner, represents the height information with RGB value, restores the actual height distribution of each point on the whole PCB bare board, then converts the complex spectrum pattern into a comprehensive deformation index, so that the warping degree of different batches has a comparable numerical scale. When the deformation index exceeds the preset threshold, not only an unqualified conclusion is given, but also the production process of the PCB is further adjusted based on the deformation index itself, realizing the leap from passive detection to active improvement, which can effectively reduce the scrap rate and improve the yield stability of batch production.

Description

Technical Field

[0001] This invention relates to the field of intelligent PCB board inspection technology, and more specifically, to a method for improving PCB board warping. Background Technology

[0002] During the production of bare PCBs (empty substrates without any soldered components), mismatches in the coefficients of thermal expansion (CTE) of materials during processes such as lamination, bonding, and cooling can easily cause warping (commonly known as "board warping"). If board warping is not effectively controlled at the bare board stage, it will directly lead to serious quality problems such as component misalignment, cold solder joints, tombstoning, and even failure to pass reflow soldering during subsequent surface mount technology (SMT).

[0003] Currently, the most common method for detecting board warping in the industry is the plug gauge method (manually measuring the height of the board corner warping), which is inefficient and greatly affected by the operator.

[0004] Therefore, there is an urgent need for a method to improve board warping that can quickly and comprehensively quantify bare board deformation and proactively optimize the production process based on the deformation. Summary of the Invention

[0005] The purpose of this invention is to provide a method for improving PCB board warping, so as to improve the above-mentioned problems.

[0006] To achieve the above objectives, the embodiments of this application provide the following technical solutions: This application provides a method for improving PCB board warping. The method includes: scanning the PCB board placed on a quality inspection platform with a laser scanner to obtain a corresponding spectral pattern, wherein the RGB values ​​in the spectral pattern represent height; analyzing the deformation index of the current PCB board based on the spectral pattern; and adjusting the PCB board manufacturing process based on the deformation index if the deformation index is greater than a preset threshold.

[0007] Optionally, analyzing the deformation index of the current PCB board based on the spectral pattern includes: Based on the spectral patterns, the overall deformation of the PCB board and the deformation of important local patch areas are analyzed sequentially, and the deformation index of the PCB board is evaluated by weighting the overall deformation and the deformation of important local patch areas.

[0008] Optionally, the overall deformation of the PCB board can be analyzed based on spectral patterns, including: The preset grid is vertically mapped onto the spectral pattern to obtain multiple horizontal RGB lines and vertical RGB lines with spectral information. Based on the RGB values ​​of the RGB horizontal lines, the maximum and minimum height values ​​of each RGB horizontal line are calculated, and thus the first height difference of each RGB horizontal line is obtained. Based on the RGB values ​​of the RGB vertical lines, the maximum and minimum height values ​​of each RGB vertical line are calculated, and then the second height difference of each RGB vertical line is obtained. The lateral deformation index and longitudinal deformation index of the PCB board are obtained by summing up the multiple first height differences and multiple second height differences respectively. The overall deformation index of the PCB board is calculated using a weighted algorithm based on the lateral deformation index and the longitudinal deformation index.

[0009] Optionally, deformation of key patch areas on the PCB board can be analyzed based on spectral patterns, including: Based on the PCB model, the coordinate information of the preset important patch area is retrieved and mapped onto the spectral pattern to obtain the local spectral pattern corresponding to the important patch grid area; The local spectral pattern is divided into multiple observation units, and the RGB value corresponding to each observation unit is identified. Then, the RGB mean square error corresponding to multiple observation units is calculated and denoted as the local deformation index of the PCB board. Then, the PCB board deformation index is obtained by weighting the overall deformation index of the PCB board and the local deformation index of the PCB board.

[0010] Optionally, the manufacturing process of the PCB board can be adjusted based on the deformation index, including: Obtain the lateral deformation index and / or longitudinal deformation index of multiple consecutive batches of PCB boards; In response to the lateral deformation index and / or longitudinal deformation index continuously exceeding the preset bare board warpage threshold within a preset statistical period, a lamination cooling rate adjustment command is generated and sent to the PCB lamination process equipment. The lamination cooling rate adjustment command is used to control the cooling stage of the lamination process after resin curing, adjusting the cooling rate from a first rate range to a second rate range lower than the first rate range, so as to reduce the difference in shrinkage stress between the substrate copper foil and the insulating dielectric layer.

[0011] The beneficial effects of this invention are as follows: This invention acquires spectral patterns using a laser scanner and represents height information with RGB values. It can restore the actual height distribution of every point on the entire bare PCB board with high spatial resolution. Then, the complex spectral patterns are transformed into a comprehensive "deformation index," which provides a comparable numerical scale for the warpage of different batches. When the deformation index exceeds a preset threshold, it not only gives a "non-conforming" conclusion but also further adjusts the PCB manufacturing process based on the deformation index itself. This achieves a leap from "passive detection" to "proactive improvement," effectively reducing scrap rates and improving the yield stability of mass production.

[0012] Secondly, this solution features fast laser scanning speed, and the spectral pattern analysis can be completed automatically by computer. The entire process of deformation index calculation and threshold comparison is performed without human intervention, making it suitable for integration into PCB automated production lines as an online monitoring module with strong adaptability.

[0013] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of a method for improving PCB board warping as described in an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0017] It should be noted that similar reference numerals or letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0018] Example 1:

[0019] like Figure 1 As shown, this embodiment provides a method for improving PCB board warping, the method including steps S100 and S200.

[0020] Step S100: Scan the PCB board placed on the quality inspection platform with a laser scanner to obtain the corresponding spectral pattern. The RGB values ​​in the spectral pattern represent the height. Step S200: Analyze the deformation index of the current PCB board based on the spectral pattern, and adjust the PCB board manufacturing process based on the deformation index if the deformation index is greater than a preset threshold.

[0021] In this embodiment, step S200 involves analyzing the deformation index of the current PCB board based on the spectral pattern and adjusting the PCB board manufacturing process based on the deformation index when the deformation index exceeds a preset threshold. The general principle behind this implementation is as follows: Based on spectral patterns, the overall deformation of the PCB board and the deformation of key local patch areas are analyzed sequentially. The deformation index of the PCB board is then weighted and evaluated based on the overall deformation and the deformation of key local patch areas. In other words, a comprehensive judgment is made by combining a rough overall deformation assessment with a precise deformation assessment of key areas (such as BGA pads, QFP pin areas, and large-size chip mounting areas). Specifically, this can be achieved as follows: Step S210: Vertically map the preset grid onto the spectral pattern to obtain multiple horizontal RGB lines and vertical RGB lines with spectral information; Step S220: Based on the RGB values ​​of the RGB horizontal lines, calculate the maximum and minimum height values ​​of each RGB horizontal line, and then obtain the first height difference of each RGB horizontal line; Step S230: Based on the RGB values ​​of the RGB vertical lines, calculate the maximum and minimum height values ​​of each RGB vertical line, and then obtain the second height difference of each RGB vertical line; Step S240: The multiple first height differences and multiple second height differences are summed to obtain the lateral deformation index and longitudinal deformation index of the PCB board. Step S250: Calculate the overall deformation index of the PCB board based on the lateral deformation index and the longitudinal deformation index using a weighted algorithm; Step S260: Based on the PCB model, retrieve the coordinate information of the preset important patch area and map it onto the spectral pattern to obtain the local spectral pattern corresponding to the important patch grid area; Step S270: Divide the local spectral pattern into multiple observation units and identify the RGB value corresponding to each observation unit. Then calculate the RGB mean square error corresponding to multiple observation units, which is denoted as the local deformation index of the PCB board. Then, based on the overall deformation index of the PCB board and the local deformation index of the PCB board, obtain the PCB board deformation index through a weighted algorithm. Step S280: When the deformation index is greater than the preset threshold, continuously acquire the lateral deformation index and / or longitudinal deformation index of multiple consecutive batches of PCB boards. If the lateral deformation index and / or longitudinal deformation index continuously exceed the preset bare board warpage threshold within a preset statistical period, a lamination cooling rate adjustment command is generated and sent to the PCB lamination process equipment. The lamination cooling rate adjustment command is used to control the cooling stage of the lamination process after resin curing, adjusting the cooling rate from a first rate range to a second rate range lower than the first rate range, so as to reduce the difference in shrinkage stress between the substrate copper foil and the insulating dielectric layer, thereby improving the bending condition of the finished bare PCB.

[0022] This embodiment does not require processing the 3D data of all pixels on the entire board surface. Instead, it uses a limited number of horizontal and vertical lines (e.g., one line every 5-10 mm) to characterize the deformation distribution of the entire board in the X and Y principal directions. This "line sampling" method significantly reduces the amount of data, making it suitable for online high-speed inspection scenarios and enabling preliminary assessment of overall deformation within milliseconds.

[0023] Secondly, to further improve data processing efficiency, in the overall deformation assessment, only the highest and lowest points are extracted, ignoring the shape parameters of the horizontal and vertical lines. The maximum and minimum height values ​​for each horizontal / vertical line are used to calculate the first / second height difference, and then the differences across all lines are summed. This "range summation" method amplifies the global trend and avoids the excessive influence of minor local bumps and dips on the overall deformation assessment.

[0024] The lateral and longitudinal deformation indices can be combined into a comprehensive overall deformation index using a weighted algorithm (e.g., weighting according to the aspect ratio of the board or the expected direction of force). Different weights can be set for square or rectangular boards, improving the method's adaptability to different PCB bare board shapes.

[0025] As can be seen, the reference elements (height difference between horizontal and vertical lines) of the overall deformation index are few, which is a relatively coarse evaluation method. The overall deformation is only based on horizontal and vertical line sampling, which may miss the local deformation peak (less than the maximum height value) between two sampling lines and on the sampling lines. Therefore, for important patch areas, this embodiment directly obtains the local spectral pattern (a continuous two-dimensional RGB height map) corresponding to the important patch area, and then divides it into multiple small observation units. This "in-plane segmentation + statistics" method can capture any micro-ripples in the area, such as a local bulge caused by a tiny bubble inside the BGA pad area, which is likely to be missed by the overall horizontal and vertical line sampling.

[0026] The impact of critical placement areas on placement quality depends not on the maximum height difference, but on the dispersion (mean square error) of the height of all points within the area. This is because during placement, all pads within the entire area must be substantially coplanar to ensure reliable contact at each solder joint. Calculating the RGB mean square error of multiple observation units as a "local deformation index" directly quantifies the flatness consistency of the area, which is more meaningful for engineering purposes than simply considering the maximum height difference.

[0027] Finally, a weighted algorithm is used to combine the overall deformation index and the local deformation index into the final PCB board deformation index. This means: Even if the overall warping of the board is not significant, if the local variance of a certain important patch area (such as the BGA area) exceeds the standard, the comprehensive index will still exceed the threshold, triggering process adjustments.

[0028] Conversely, if the overall board warps significantly but the important patch area happens to be located on a relatively flat "neutral surface", the local deformation index may be low, and the overall index may not exceed the threshold, thus avoiding excessive process intervention.

[0029] This weighting mechanism ensures that the deformation index truly reflects the actual failure risk in surface mount technology (SMT) production, rather than simply pursuing board-level flatness.

[0030] This embodiment acquires spectral patterns using a laser scanner and represents height information using RGB values. It can restore the actual height distribution of every point on the entire bare PCB board with high spatial resolution. Then, the complex spectral patterns are converted into a comprehensive "deformation index," which provides a comparable numerical scale for the degree of warpage in different batches. When the deformation index exceeds a preset threshold, it not only gives a "non-conforming" conclusion but also further adjusts the PCB manufacturing process based on the deformation index itself. This achieves a leap from "passive detection" to "proactive improvement," which can effectively reduce the scrap rate and improve the yield stability of mass production.

[0031] Secondly, this solution features fast laser scanning speed, and the spectral pattern analysis can be completed automatically by computer. The entire process of deformation index calculation and threshold comparison is performed without human intervention, making it suitable for integration into PCB automated production lines as an online monitoring module with strong adaptability.

[0032] Example 2:

[0033] This embodiment, based on Embodiment 1, further supplements the explanation of the implementation details of calculating the maximum and minimum height values ​​of each RGB horizontal line based on the RGB values ​​of the RGB horizontal lines described in step S220. Step S230 is similar and will not be repeated here. Specifically, it is as follows: Step S221: Based on the changes in the RGB values ​​of the RGB horizontal line, determine whether there are any RGB abrupt change points on the RGB horizontal line. The RGB abrupt change point is a point where the slope of the RGB change is greater than a preset slope threshold. If it exists, extend a preset length to both sides with the mutation point as the center line, and mark the local line segments that are extended to the RGB horizontal line as invalid line segments. The invalid line segments do not participate in the selection range of the maximum and minimum height values.

[0034] That is, when the RGB horizontal line passes through the through hole on the PCB board, the sampling influence of the through hole segment on the maximum and minimum values ​​needs to be eliminated.

[0035] Example 3:

[0036] This embodiment, based on Embodiment 1, details the specific implementation of step S260, which involves retrieving the coordinate information of a preset important patch area based on the PCB model and mapping it onto the spectral pattern to obtain the local spectral pattern corresponding to the important patch grid area. Step S261: Retrieve the coordinate information of the preset important patch area based on the PCB model. The coordinate information is a two-dimensional plane reference coordinate set of the important patch grid area in the flat state of the PCB board. Step S262: On the spectral pattern, determine a deformation evaluation buffer that covers the two-dimensional plane reference coordinate set. The determination method is as follows: construct a rectangle that is closest to the PCB board in the flat state based on the outline of the spectral pattern. Then, enlarge the area corresponding to the two-dimensional plane reference coordinate set by a certain proportion to obtain the deformation evaluation buffer. The enlargement operation is to enlarge the area proportionally with the midpoint of the area corresponding to the two-dimensional plane reference coordinate set as the fixed point. Step S263: Vertically map the deformation evaluation buffer to the spectral pattern to obtain the local spectral pattern corresponding to the important patch mesh area.

[0037] When a PCB bare board warps, the actual spatial position of each point on its surface will undergo in-plane displacement relative to the theoretical two-dimensional coordinates in its flat state (for example, when a corner of the board warps, the area shrinks or stretches towards the center or edge of the board). If the spectral pattern is directly cropped according to the theoretical coordinate range, the actual important patch features may partially shift out of the cropped area, resulting in the acquisition of local spectral patterns lacking key deformation information. This embodiment constructs a proportionally enlarged deformation evaluation buffer (e.g., enlarged by 10% to 30%) to ensure that regardless of the overall direction of the important patch area's shift, its actual deformation influence range can still be completely contained within the buffer. This fundamentally solves the technical problem of "inaccurate theoretical coordinate cropping" and avoids misjudging the degree of local deformation due to cropping errors.

[0038] Secondly, the zoom-in operation, which uses the midpoint of the region corresponding to the two-dimensional reference coordinate set as a fixed point for proportional zoom-in, has the following advantages: Symmetrical expansion: The enlarged buffer still expands uniformly outward from the center of the original theoretical region. This symmetry ensures that deformation analysis within the buffer does not introduce additional directional bias, facilitating statistical unbiasedness in subsequent calculations of the RGB mean square error.

[0039] Preservation of relative deformation relationships: Because the magnification is proportional, the relative positional relationship between points within the original theoretical region and the edge of the buffer zone is preserved. When warping causes different points within the region to have different displacements (e.g., larger displacement at the edge and smaller displacement at the center), the magnified buffer zone can still cover these points according to the original geometric proportions without destroying the spatial gradient characteristics of the deformation.

[0040] Secondly, a reference rectangle is constructed based on the spectral pattern outline to achieve robustness of coordinate alignment. In step S262, a rectangle is first constructed based on the spectral pattern outline that is closest to and the same size as the PCB board in its flat state. The purpose of this operation is: Eliminating the effects of overall PCB translation and rotation: During actual scanning, the placement and angle of the bare PCB board on the quality inspection platform may deviate slightly from the preset direction of the theoretical coordinate system. By extracting the actual outline of the PCB board from the spectral pattern and constructing the minimum bounding rectangle, a reference frame with the same theoretical board size but aligned with the measured data in position / orientation can be obtained, thus accurately mapping the theoretical coordinate set onto the measured spectral pattern.

[0041] Because the deformation assessment buffer fully covers the actual deformation range of the important patch area, the subsequent local spectral pattern obtained by vertically mapping the buffer contains both points within the original theoretical region and information about the surrounding transition zones that have shifted into the buffer due to warping. This allows: The calculation of local RGB mean square error (local deformation index) is more accurate: it will not underestimate the mean square error because it misses the high or low points of the actual deformation.

[0042] To avoid "false acceptance" judgments: If the selected area happens to avoid a local protrusion due to coordinate offset, and this protrusion is not included in the standard deviation, the PCB may be mistakenly judged as acceptable. This embodiment uses a buffer design to ensure that any deformation features located near the theoretical position of important surface mount areas are included in the evaluation, thereby significantly reducing the risk of missed detections.

[0043] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for improving PCB board warping, characterized in that, The method includes: The PCB board placed on the quality inspection platform is scanned by a laser scanner to obtain the corresponding spectral pattern, and the RGB values ​​in the spectral pattern represent the height. The deformation index of the current PCB board is analyzed based on the spectral pattern, and the manufacturing process of the PCB board is adjusted based on the deformation index if the deformation index is greater than a preset threshold. The analysis of the deformation index of the current PCB board based on the spectral pattern includes: Based on the spectral patterns, the overall deformation of the PCB board and the deformation of important local patch areas are analyzed sequentially, and the deformation index of the PCB board is evaluated by weighting the overall deformation and the deformation of important local patch areas. Among them, the analysis of the overall deformation of the PCB board based on spectral patterns includes: The preset grid is vertically mapped onto the spectral pattern to obtain multiple horizontal RGB lines and vertical RGB lines with spectral information. Based on the RGB values ​​of the RGB horizontal lines, the maximum and minimum height values ​​of each RGB horizontal line are calculated, and thus the first height difference of each RGB horizontal line is obtained. Based on the RGB values ​​of the RGB vertical lines, the maximum and minimum height values ​​of each RGB vertical line are calculated, and then the second height difference of each RGB vertical line is obtained. The lateral deformation index and longitudinal deformation index of the PCB board are obtained by summing up the multiple first height differences and multiple second height differences respectively. The overall deformation index of the PCB board is calculated using a weighted algorithm based on the lateral deformation index and the longitudinal deformation index.

2. The method for improving PCB board warping according to claim 1, characterized in that, Deformation of key patch areas on PCB boards based on spectral pattern analysis, including: Based on the PCB model, the coordinate information of the preset important patch area is retrieved and mapped onto the spectral pattern to obtain the local spectral pattern corresponding to the important patch grid area; The local spectral pattern is divided into multiple observation units, and the RGB value corresponding to each observation unit is identified. Then, the RGB mean square error corresponding to multiple observation units is calculated and denoted as the local deformation index of the PCB board. Then, the PCB board deformation index is obtained by weighting the overall deformation index of the PCB board and the local deformation index of the PCB board.

3. The method for improving PCB board warping according to claim 2, characterized in that, Adjusting PCB manufacturing processes based on deformation index includes: Continuously obtain the lateral deformation index and / or longitudinal deformation index of multiple consecutive batches of PCB boards; In response to the lateral deformation index and / or longitudinal deformation index continuously exceeding the preset bare board warpage threshold within a preset statistical period, a lamination cooling rate adjustment command is generated and sent to the PCB lamination process equipment. The lamination cooling rate adjustment command is used to control the cooling stage of the lamination process after resin curing, adjusting the cooling rate from a first rate range to a second rate range lower than the first rate range, so as to reduce the difference in shrinkage stress between the substrate copper foil and the insulating dielectric layer.

Images