Method for improving dry film pressing effect and etching quality of non-adhesive layering area of PCB (Printed Circuit Board)

Abstract

The invention relates to the field of printed circuit board (PCB) manufacturing, provides a method for improving the dry film pressing effect and etching quality of a non-adhesive layered area of a PCB, and is suitable for PCB production links such as pattern electroplating, dry film pressing and etching. The method comprises the following steps: designing a film used for pattern electroplating, and changing a linear design of the edge of a non-adhesive area into an inverted raised grain or zigzag design so as to disperse and release stress generated during dry film pressing and improve the coverage and the fitting tightness of the dry film in a step area; after the outer-layer base material is pressed and before the dry film is pressed, the thickness of the bottom copper is uniformly reduced, the total amount of copper needing to be etched subsequently is reduced, and the lateral erosion risk caused by the fact that the dry film is not pressed compactly is reduced; the total thickness of the bottom copper and the surface copper of the base material is reduced to 15-20 microns, the total copper height after pattern electroplating is reduced, the absolute offset height between the first electricity and the second electricity is reduced, and the dry film is easy to compact and press-fit without bubbles.

Description

Technical Field

[0001] This application relates to the field of printed circuit board (PCB) manufacturing technology, and in particular to a method for improving the dry film bonding effect and etching quality in glue-free delamination areas of PCBs. Background Technology

[0002] In the field of printed circuit board (PCB) manufacturing, the stability of the dry film lamination process directly affects etching quality and product yield. In existing technologies, steps in adhesive-free delamination areas (such as the area separating electrical components) often lead to the following problems due to incomplete dry film lamination: 1. Dry film bonding defects: Traditional film (photomask) uses a straight line design at the edge of the adhesive-free area. Stress concentration during lamination can easily lead to wrinkles and bubbles in the dry film. Especially when the step height is large, the dry film is difficult to tightly cover the discontinuity area, resulting in poor lamination. 2. Low etching efficiency: Conventional processes do not include a copper reduction process after laminating the outer substrate. The bottom copper thickness is relatively thick, resulting in a large total amount of copper layer to be removed during etching. This leads to a long etching time and a high risk of side etching, which can easily cause incomplete etching in the two electrical regions. 3. Insufficient control of discontinuity: The total thickness of the base copper and the surface copper of the substrate was not reduced in a targeted manner, resulting in a large copper height after pattern electroplating, which led to a significant discontinuity between the first and second electroplating processes, further exacerbating the difficulty of dry film lamination.

[0003] Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the Invention

[0004] This application provides a method to improve the dry film lamination effect and etching quality in glue-free delamination areas of PCBs. It aims to solve the problem that some solutions in the prior art attempt to improve the problem through single process adjustments (such as local optimization of film patterns or individual reduction of copper thickness), but have not formed a systematic solution.

[0005] In a first aspect, embodiments of this application provide a method for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB, the method comprising: The film used for graphic electroplating was designed by changing the straight lines at the edges of the non-adhesive areas to inverted wavy or sawtooth patterns to disperse and release the stress generated during dry film lamination, thereby improving the coverage and adhesion tightness of the dry film in the stepped areas. After the outer substrate is laminated and before the dry film is pressed, the thickness of the bottom copper is uniformly reduced to reduce the total amount of copper that needs to be etched away later, thereby reducing the risk of side etching due to insufficient dry film pressing. The total thickness of the base copper and surface copper of the substrate is reduced to 15-20 micrometers, which reduces the total copper height after pattern electroplating and reduces the absolute height difference between the first and second electroplating processes, making it easier to press the dry film tightly and without bubbles.

[0006] In some embodiments, the film used for the design pattern electroplating changes the straight line design of the edge of the adhesive-free area to an inverted wavy or sawtooth design to disperse and release the stress generated during dry film lamination. This includes: simulating the stress distribution during the dry film lamination process using a finite element analysis algorithm, and generating an inverted wavy or sawtooth profile of the edge of the adhesive-free area based on the simulation results; wherein the wave crest spacing, wave depth, or sawtooth angle of the wavy pattern is automatically calculated and determined according to the substrate thickness and the copper layer discontinuity height to maximize stress dispersion.

[0007] In some embodiments, the process of uniformly thinning the bottom copper thickness after laminating the outer substrate and before drying the film includes: acquiring copper layer thickness data on the substrate surface using an online copper thickness detection sensor, calculating the micro-etching time or copper stripping depth of the copper reduction process using a closed-loop control algorithm, and establishing a mapping relationship between the amount of bottom copper thickness reduction and the difficulty of subsequent etching.

[0008] In some embodiments, reducing the total thickness of the base copper and surface copper of the substrate to 15-20 micrometers and reducing the total copper height after pattern electroplating includes: establishing a mathematical model of the total copper thickness and the difference height based on historical data of the first and second electroplating breaks using a regression analysis algorithm; adjusting the thickness ratio of the base copper and surface copper according to the optimal solution output by the model, so that the total thickness is controlled within the range of 15-20 micrometers, while meeting the conductivity requirements.

[0009] In some embodiments, the method further includes: during the film design stage, using machine learning algorithms to train a dry film wrinkle prediction model, inputting parameters such as the shape of the glue-free area and the copper layer discontinuity height, and outputting the optimal structural parameters for the inverted wave pattern or sawtooth design, so as to predict and reduce stress concentration during pressing in advance.

[0010] In some embodiments, the method further includes: in the copper reduction process, acquiring images of the substrate surface in real time through a machine vision system, using an image recognition algorithm to detect the uniformity of the copper layer after thinning, and if a local thickness abnormality is detected, automatically adjusting the spray pressure of the micro-etching equipment or the travel speed of the copper planing tool to achieve dynamic compensation.

[0011] In some embodiments, the method further includes: establishing a process database of dry film lamination effect and copper thickness structure, analyzing historical yield data through data mining algorithms, generating a copper thickness reduction strategy library for different break heights, and automatically matching the optimal combination of bottom copper and surface copper thicknesses during new batch production.

[0012] In some embodiments, the method further includes: after pressing the dry film, using infrared thermal imaging technology to detect the dry film bonding temperature distribution in the step area, calculating the stress value of the temperature abnormal area through a heat conduction model algorithm, and automatically adjusting the temperature or pressure parameters of the pressing equipment for secondary pressing if the stress value exceeds a preset threshold.

[0013] In some embodiments, the method further includes: predicting the probability of lateral etching caused by poor dry film pressing using a neural network algorithm before the etching process, and dynamically adjusting the etching solution concentration or etching time based on the prediction results to form intelligent linkage control of pressing effect and etching parameters.

[0014] In some embodiments, the method further includes: using a multi-objective optimization algorithm to collaboratively optimize the film pattern design parameters, copper thickness reduction, and total copper thickness, with dry film bonding tightness, etching yield, and production efficiency as optimization objectives, to generate a globally optimal combination of process parameters.

[0015] This application effectively disperses bonding stress and reduces dry film wrinkles and bubbles through a wavy or sawtooth film design, thereby improving the dry film coverage density in stepped areas and solving the problem of stress concentration at traditional straight edges. The copper reduction process reduces the total copper thickness of subsequent etching, shortens etching time, and reduces side etching. Combined with a thinner copper structure (15-20 micrometers), the height difference between the first and second electrical connections is reduced, the difficulty of dry film bonding is reduced, and the defect rate of incomplete etching is significantly reduced. Through the synergistic effect of the three processes, the electrical test yield is improved, and a systematic solution is achieved for the bonding and etching problems in glue-free delamination areas of PCBs, reducing production costs and improving production efficiency.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart illustrating the steps of a method for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB, according to an embodiment of this application. Figure 2 This is a schematic block diagram of a system for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB, provided in one embodiment of this application. Figure 3 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0022] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0023] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0024] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] In the field of printed circuit board (PCB) manufacturing, the stability of the dry film lamination process directly affects etching quality and product yield. In existing technologies, steps in adhesive-free delamination areas (such as the area separating electrical components) often lead to the following problems due to incomplete dry film lamination: 1. Dry film bonding defects: Traditional film (photomask) uses a straight line design at the edge of the adhesive-free area. Stress concentration during lamination can easily lead to wrinkles and bubbles in the dry film. Especially when the step height is large, the dry film is difficult to tightly cover the discontinuity area, resulting in poor lamination. 2. Low etching efficiency: Conventional processes do not include a copper reduction process after laminating the outer substrate. The bottom copper thickness is relatively thick, resulting in a large total amount of copper layer to be removed during etching. This leads to a long etching time and a high risk of side etching, which can easily cause incomplete etching in the two electrical regions. 3. Insufficient control of discontinuity: The total thickness of the base copper and the surface copper of the substrate was not reduced in a targeted manner, resulting in a large copper height after pattern electroplating, which led to a significant discontinuity between the first and second electroplating processes, further exacerbating the difficulty of dry film lamination.

[0026] In existing technologies, some solutions attempt to improve the problem through single process adjustments (such as local optimization of film patterns or separate reduction of copper thickness), but no systematic solution has been formed. For example, simply modifying the film pattern without combining it with copper thickness structure adjustment cannot fundamentally reduce the impact of the break height on lamination; adding a copper reduction process alone without matching film design optimization makes it difficult to eliminate dry film wrinkles caused by stress concentration; traditional processes do not combine the three improvement measures, and cannot achieve the synergistic effect of "stress dispersion - copper thickness optimization - break height reduction", resulting in a persistently high defect rate.

[0027] To solve the above problem, please refer to Figure 1 This application provides a method for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB, applicable to computer equipment. The computer equipment can be deployed on a single server or server cluster. It can also be deployed on handheld terminals, laptops, wearable devices, or robots, etc. It should be noted that all information involved in the method provided in this application is extracted with the authorization of the relevant user and in accordance with relevant regulations, and will not infringe on user privacy.

[0028] The provided method for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB includes steps S101 to S103. Details are as follows: Step S101. Design the film used for pattern electroplating, and change the straight line design of the edge of the non-adhesive area to an inverted wave pattern or a sawtooth design to disperse and release the stress generated during dry film lamination, and improve the coverage and adhesion tightness of the dry film in the step area.

[0029] Specifically, by modifying the edge shape of the adhesive-free area of ​​the film (photomask) used for pattern electroplating, the traditional straight line design is replaced with an inverted wave pattern or a sawtooth structure to change the stress distribution path during dry film lamination, avoid wrinkles and bubbles caused by stress concentration, and improve the dry film adhesion tightness in the stepped area.

[0030] Computer-aided design (CAD) modeling uses PCB design software to create film graphics. For the edges of non-adhesive areas (such as the boundaries of plated areas), the software's drawing tools modify straight edges into continuous inverted wavy or serrated outlines.

[0031] The crest spacing can be set to 50-100 micrometers, and the wave depth is 20-50 micrometers; the sawtooth angle can be set to 60°-120°, and the tooth pitch is matched with the crest spacing. The specific values ​​are optimized through computer simulation (such as finite element analysis algorithm), with the substrate thickness and copper layer discontinuity height as input parameters, and the output is the geometric structure with optimal stress dispersion.

[0032] Import the designed film image into the stress simulation system, and use computer algorithms to simulate the stress distribution during the dry film pressing process, marking areas of high stress concentration (such as the corners of steps).

[0033] The density, angle, or depth of the wavy / sawtooth pattern are dynamically adjusted based on the simulation results until the uniformity of stress distribution is improved to a preset threshold (e.g., the stress standard deviation is reduced by more than 30%).

[0034] Film production and verification involves transmitting optimized graphic data to a laser plotter, which then etches inverted wavy or serrated edges onto the film to ensure line accuracy of ±5 micrometers.

[0035] The film is scanned by optical inspection equipment (such as a film inspector), and the computer automatically compares the design file with the actual drawing to verify the consistency of the edge shape and avoid stress dispersion failure caused by processing errors.

[0036] Step S102. After the outer substrate is laminated and before the dry film is pressed, the thickness of the bottom copper is uniformly reduced to reduce the total amount of copper that needs to be etched away later, thereby reducing the risk of side etching caused by the dry film not being properly pressed.

[0037] Specifically, after the outer substrate is laminated and before the film is dried, an additional copper reduction process (surface micro-etching or copper stripping) is added to uniformly reduce the thickness of the bottom copper, reduce the total amount of copper layer to be removed in subsequent etching processes, shorten etching time and reduce the risk of side etching, while also reducing the initial height difference in the step area.

[0038] Copper reduction process options include: Chemical etching: The laminated substrate is immersed in an acidic etching solution (such as sodium persulfate solution). By controlling the etching time (usually 5-10 minutes) and solution concentration (20-30 g / L), the copper surface is uniformly etched, and the thickness reduction is controlled at 3-5 micrometers. Physical copper planing: A CNC copper planing machine is used to physically grind the copper surface by rotating cutters or abrasive belts. The planing depth is automatically adjusted by a computer program according to a preset thinning amount (such as 8 micrometers) to adjust the cutter speed and pressure, ensuring a surface roughness Ra ≤ 1.5 micrometers.

[0039] Online detection and closed-loop control involve installing non-contact copper thickness sensors (such as X-ray thickness gauges or eddy current thickness gauges) at the copper reduction equipment outlet to collect real-time copper thickness data on the substrate surface. The computer system compares the measured values ​​with the target values ​​(such as a copper thickness of 3-5 micrometers after thinning). If a thickness deviation exceeding ±5% is detected, the system automatically adjusts the micro-etching fluid flow rate, temperature, or copper planing machine tool parameters, forming a "detection-calculation-adjustment" closed-loop control to ensure uniform copper thickness across the entire board.

[0040] After copper reduction is completed, the substrate immediately enters the washing and drying process. The computer controls the transmission speed (e.g., 2-3 meters / minute) to avoid copper surface oxidation affecting the adhesion of subsequent dry film lamination.

[0041] Step S103. Reduce the total thickness of the bottom copper and top copper of the substrate to 15-20 micrometers, reduce the total copper height after pattern electroplating, reduce the absolute difference height between the first and second electroplating, and make the dry film easy to press tightly and without bubbles.

[0042] Specifically, the total thickness of the base copper and surface copper of the substrate is reduced from the conventional standard to 15-20 micrometers, thereby reducing the total copper height after pattern electroplating. This reduces the absolute height difference between the first and second electroplating processes (e.g., from 30 micrometers to below 15 micrometers), allowing the dry film to more tightly cover the step area during lamination and reducing bubble formation.

[0043] During the substrate lamination stage, the computer selects the base copper thickness (such as 10-micron or 12-micron electrolytic copper foil) from a predefined copper foil thickness library according to the product design requirements. It controls the bonding force between the copper foil and the substrate through lamination parameters (temperature 180-200℃, pressure 3-5 MPa) to ensure a peel strength ≥1.5 N / mm.

[0044] The thickness of the surface copper plating is controlled by precisely calculating the plating time using an ampere-hour meter during the pattern plating process. For example, if the base copper is 10 micrometers and the target total thickness is 18 micrometers, the surface copper plating thickness is controlled to be 8 micrometers, the plating current density is set to 20-25 amperes per square decimeter, and the plating time is monitored in real time by a computer and automatically terminated.

[0045] The step height simulation and verification utilizes computer 3D modeling software to simulate the step height of the first and second electrical regions based on the thickness data of the bottom copper and the surface copper (formula: step height = total copper thickness - copper thickness retained after etching).

[0046] If the simulated discontinuity height exceeds 15 micrometers, the system will automatically alarm and prompt adjustments to the copper thickness ratio (e.g., increase the bottom copper thickness to 12 micrometers and reduce the surface copper thickness to 6 micrometers) until the discontinuity meets the process requirements (≤10 micrometers).

[0047] Copper thickness data, film design, and copper reduction process parameters are shared in real time through the Manufacturing Execution System (MES). For example, after the total copper thickness is reduced, the film wavy pattern parameters are automatically adapted to a denser sawtooth design to cope with stress distribution changes under smaller breaks.

[0048] In some embodiments, the film used for the design pattern electroplating changes the straight line design of the edge of the adhesive-free area to an inverted wavy or sawtooth design to disperse and release the stress generated during dry film lamination. This includes: simulating the stress distribution during the dry film lamination process using a finite element analysis algorithm, and generating an inverted wavy or sawtooth profile of the edge of the adhesive-free area based on the simulation results; wherein the wave crest spacing, wave depth, or sawtooth angle of the wavy pattern is automatically calculated and determined according to the substrate thickness and the copper layer discontinuity height to maximize stress dispersion.

[0049] By introducing finite element analysis algorithms into the film design, the stress distribution of dry film bonding is simulated, and the inverted wavy or sawtooth contour parameters of the edge of the glue-free area are dynamically optimized to maximize stress dispersion.

[0050] The finite element model is constructed using simulation software to create a three-dimensional finite element model of the dry film-copper layer-substrate, defining material properties (such as the elastic modulus of the dry film and the Poisson's ratio of the copper layer). The lamination process parameters (temperature 120-150℃, pressure 0.8-1.2MPa) are input to simulate the stress transmission path during lamination, marking high stress concentration points at the edges of the adhesive-free areas (such as areas with stress values ​​> 5MPa).

[0051] The wavy pattern parameters are automatically generated based on the substrate thickness (e.g., 1.0mm or 1.6mm) and the copper layer discontinuity height (e.g., 10-20 micrometers). Through iterative calculation, the wavy pattern geometry parameters are generated: when the discontinuity is >15 micrometers, the crest spacing is automatically set to 80 micrometers, the wave depth to 40 micrometers, and the sawtooth angle to 90°; when the discontinuity is ≤15 micrometers, the crest spacing is adjusted to 50 micrometers, the wave depth to 20 micrometers, and the sawtooth angle to 120°. The output film graphic file is automatically imported into a laser plotter, ensuring an edge contour accuracy of ±3 micrometers.

[0052] Simulation verification and iteration involves performing secondary stress simulation on the generated film design. If the stress concentration factor (maximum stress / average stress) is greater than 1.8, the wavy density is automatically adjusted (increasing the number of peaks by 20%) until the factor is less than or equal to 1.5.

[0053] In some embodiments, the process of uniformly thinning the bottom copper thickness after laminating the outer substrate and before drying the film includes: acquiring copper layer thickness data on the substrate surface using an online copper thickness detection sensor, calculating the micro-etching time or copper stripping depth of the copper reduction process using a closed-loop control algorithm, and establishing a mapping relationship between the amount of bottom copper thickness reduction and the difficulty of subsequent etching.

[0054] By using online copper thickness detection and closed-loop control algorithms, a dynamic mapping between copper reduction and etching difficulty is established, enabling precise control of the bottom copper thickness.

[0055] Real-time data acquisition is achieved by installing an X-ray thickness gauge at the inlet of the copper reduction equipment, scanning the substrate surface every 5 seconds to obtain copper thickness data at 200×200 points across the entire board, and generating a thickness thermal map (resolution 0.1 micrometers).

[0056] The closed-loop control algorithm employs a PID control algorithm to calculate the target thinning amount (e.g., target base copper thickness = original thickness - 8 micrometers). Based on the deviation between the measured thickness and the target value, it dynamically adjusts the following: chemical micro-etching: micro-etching solution temperature (accuracy ±0.5℃), spray flow rate (accuracy ±2%); physical copper planing: tool speed (accuracy ±50rpm), feed rate (accuracy ±0.1mm / s). An etching difficulty mapping table is established: if the subsequent etched copper thickness is >15 micrometers, the thinning amount automatically increases by 2 micrometers; if ≤15 micrometers, the thinning amount maintains the standard value. The algorithm is trained using historical data to ensure that the etching time prediction error is ≤5%.

[0057] If a local copper thickness deviation > ±10% is detected, the system will automatically trigger substrate marking and perform manual etching or planing before pressing the dry film to ensure uniform thickness of the entire board.

[0058] In some embodiments, reducing the total thickness of the base copper and surface copper of the substrate to 15-20 micrometers and reducing the total copper height after pattern electroplating includes: establishing a mathematical model of the total copper thickness and the difference height based on historical data of the first and second electroplating breaks using a regression analysis algorithm; adjusting the thickness ratio of the base copper and surface copper according to the optimal solution output by the model, so that the total thickness is controlled within the range of 15-20 micrometers, while meeting the conductivity requirements.

[0059] Based on historical fault data, a mathematical model of the total copper thickness and fault height is established through regression analysis algorithm to optimize the ratio of bottom copper to top copper thickness and balance conductivity and fault control.

[0060] Historical production data were collected, with variables including bottom copper thickness (X1), top copper thickness (X2), and drop height (Y). A multiple linear regression model was fitted using the least squares method: Y = 0.8X1 + 1.2X2 - 5 (unit: micrometers). The model passed the F-test (significance level α = 0.05), and the coefficient of determination R² > 0.92, validating the model's effectiveness.

[0061] Thickness ratio optimization involves inputting the target drop height (e.g., ≤10 micrometers) and using a model to inversely calculate the allowable total copper thickness range (X1+X2≤20 micrometers), combined with conductivity requirements (copper layer resistivity ≤1.8×10⁻⁶). -8(Ω·m), with priority given to a bottom copper thickness ≥10 micrometers (to improve signal integrity) and a top copper thickness ≤10 micrometers (to reduce electroplating costs).

[0062] The process parameters are automatically synchronized to the lamination process (bottom copper foil selection) and the electroplating process (surface copper thickness setting) through the optimal ratio output by the model, and a process work order QR code is generated to realize cross-process traceability.

[0063] In some embodiments, the method further includes: during the film design stage, using machine learning algorithms to train a dry film wrinkle prediction model, inputting parameters such as the shape of the glue-free area and the copper layer discontinuity height, and outputting the optimal structural parameters for the inverted wave pattern or sawtooth design, so as to predict and reduce stress concentration during pressing in advance.

[0064] By using machine learning algorithms to train a dry film wrinkle prediction model, stress concentration risks can be predicted during the film design stage, and the parameters of the corrugated structure can be optimized in advance.

[0065] Model training was conducted by collecting film design parameters (wavy spacing, wave depth, and sawtooth angle) and wrinkle rate data after pressing. A prediction model was constructed using the random forest algorithm, and the feature importance was ranked as follows: discontinuity height (40%) > area of ​​glue-free region (30%) > wavy spacing (20%).

[0066] The model's wrinkle rate prediction error on the test set is ≤8%, and optimization is triggered when the wrinkle rate is >15%.

[0067] Real-time prediction and adjustment are achieved by embedding a model API in the film design software. The shape of the glue-free area of ​​the current product (such as a polygon or circle) and the break height (obtained through the model in Example 3) are input. The model outputs suggested wavy parameters in real time: if the predicted wrinkle rate is >20%, the wavy spacing is automatically reduced from 50 micrometers to 30 micrometers and the wave depth is increased by 10 micrometers; if the predicted wrinkle rate is ≤10%, the default parameters are maintained to improve design efficiency.

[0068] Model iteration updates are achieved by automatically importing the latest production data (including new process parameters) weekly and retraining the model to ensure adaptability to new PCB products.

[0069] In some embodiments, the method further includes: in the copper reduction process, acquiring images of the substrate surface in real time through a machine vision system, using an image recognition algorithm to detect the uniformity of the copper layer after thinning, and if a local thickness abnormality is detected, automatically adjusting the spray pressure of the micro-etching equipment or the travel speed of the copper planing tool to achieve dynamic compensation.

[0070] By using machine vision and image recognition algorithms, the uniformity of the copper layer after copper reduction is monitored in real time, and the micro-etching / copper-cutting parameters are dynamically adjusted to achieve automatic compensation for local thickness anomalies.

[0071] Image acquisition and processing involves installing a linear CCD camera at the copper reduction equipment outlet to acquire images of the substrate surface at a resolution of 2000 dpi. The copper layer area is extracted using a threshold segmentation algorithm, and the standard deviation of grayscale values ​​(reflecting thickness uniformity) is calculated.

[0072] Set a uniformity threshold: when the grayscale standard deviation is >15, it is judged as a local anomaly (such as edge over-etching or center under-etching).

[0073] Dynamic compensation strategy: If the thickness of the edge area is detected to be too thick (low gray value), the flow rate of the micro-etching liquid in the edge spray area will be automatically increased by 20%, or the copper planer will be slowed down by 15% in the edge area; if the thickness of the center area is too thin, the flow rate in the center spray area will be reduced by 10%, or the feed speed of the copper planer in the center area will be increased by 10%.

[0074] After compensation, the image is acquired again. If the standard deviation is ≤10, it passes; otherwise, a manual intervention process is triggered, and the anomaly type is recorded for subsequent process improvement.

[0075] In some embodiments, the method further includes: establishing a process database of dry film lamination effect and copper thickness structure, analyzing historical yield data through data mining algorithms, generating a copper thickness reduction strategy library for different break heights, and automatically matching the optimal combination of bottom copper and surface copper thicknesses during new batch production.

[0076] A process database and data mining algorithm are established to intelligently match copper thickness reduction strategies based on the height difference, enabling batch reuse and optimization of process parameters.

[0077] The database stores fields including: product model, break height, bottom copper thickness, top copper thickness, yield, etching time, etc. It uses MySQL database management and supports data storage of millions of records and second-level query.

[0078] The data mining algorithm uses K-means clustering to group the height of the gap (e.g., <10 micrometers, 3-5 micrometers, >15 micrometers). For each group, association rule analysis is performed to extract the copper thickness strategy corresponding to high yield (>85%): gap <10 micrometers: bottom copper + surface copper = 15 micrometers (e.g., 10+5); gap 3-5 micrometers: bottom copper + surface copper = 18 micrometers (e.g., 12+6); gap >15 micrometers: bottom copper + surface copper = 20 micrometers (e.g., 15+5).

[0079] When producing a new batch, the system inputs the predicted value of the break height, automatically matches the optimal copper thickness combination from the strategy library, supports manual fine-tuning and records the reasons for the adjustment, forming a closed loop of "data-strategy-verification".

[0080] In some embodiments, the method further includes: after pressing the dry film, using infrared thermal imaging technology to detect the dry film bonding temperature distribution in the step area, calculating the stress value of the temperature abnormal area through a heat conduction model algorithm, and automatically adjusting the temperature or pressure parameters of the pressing equipment for secondary pressing if the stress value exceeds a preset threshold.

[0081] Infrared thermal imaging and heat conduction models are used to detect areas of abnormal temperature during dry film bonding. Stress concentration is eliminated through secondary pressing, thereby improving the bonding tightness.

[0082] Temperature field detection: After the film is pressed dry, the substrate surface is scanned with an infrared thermal imager at a speed of 50 frames / second with a resolution of 0.1℃ to generate a temperature distribution cloud map and mark the temperature abnormal areas (such as step areas with a temperature difference > 5℃).

[0083] Stress values ​​are calculated based on the Fourier heat conduction equation, establishing a temperature-stress mapping model: σ = k·ΔT, where k is the material's thermal stress coefficient (0.02 MPa / ℃ for dry film). Stress values ​​in abnormal areas are then calculated. When the stress value > 3 MPa, it is identified as a high-risk area, triggering a secondary pressing process.

[0084] The secondary pressing control automatically locates areas of abnormal temperature and controls the pressing equipment to apply local pressure (10% higher than the standard pressure) to these areas, maintaining the temperature at 130℃ for 5 seconds to eliminate bubbles and wrinkles. After pressing, the temperature uniformity is checked again until the temperature difference is ≤2℃.

[0085] In some embodiments, the method further includes: predicting the probability of lateral etching caused by poor dry film pressing using a neural network algorithm before the etching process, and dynamically adjusting the etching solution concentration or etching time based on the prediction results to form intelligent linkage control of pressing effect and etching parameters.

[0086] By predicting the risk of lateral etching using neural network algorithms and dynamically adjusting etching parameters, an intelligent linkage mechanism of "pressing effect - etching control" is established.

[0087] The risk prediction model is constructed using a three-layer neural network (5 input nodes, 10 hidden nodes, and 1 output node). Input parameters include: dry film wrinkling rate, copper thickness reduction, fracture height, and initial etchant concentration. The output is the probability of lateral etching risk (0-100%). The model is trained using backpropagation with mean squared error as the loss function, achieving a prediction accuracy >88% on the validation set.

[0088] Dynamic adjustment of etching parameters: If the predicted risk probability is >60%, automatically execute the following: increase the etching solution concentration from 5% to 6%; extend the etching time from 120 seconds to 150 seconds; and increase the spray pressure by 15% to enhance the etching effect. If the risk probability is ≤30%, reduce the etching time by 10% to reduce energy consumption and cost.

[0089] The effect feedback mechanism detects the amount of lateral etching after etching (e.g., using a metallographic microscope) and feeds the actual value back to the model. The model weights are automatically updated every 100 batches to improve prediction accuracy.

[0090] In some embodiments, the method further includes: using a multi-objective optimization algorithm to collaboratively optimize the film pattern design parameters, copper thickness reduction, and total copper thickness, with dry film bonding tightness, etching yield, and production efficiency as optimization objectives, to generate a globally optimal combination of process parameters.

[0091] By employing a multi-objective optimization algorithm, the film design, copper thickness reduction, and total copper thickness are optimized in a coordinated manner to balance yield, efficiency, and cost.

[0092] The optimization objectives and variables include: Objective function: Maximize dry film bonding tightness (expressed as the reciprocal of wrinkle rate); Maximize etching yield; Minimize production time (total time for film design + copper reduction + electroplating). Decision variables: Wavy pattern spacing (30-100 micrometers), thinning amount (3-5 micrometers), total copper thickness (15-20 micrometers).

[0093] The algorithm implementation uses the NSGA-II algorithm for Pareto optimization, with a population size of 100 and an iteration count of 50, to generate a set of non-dominated solutions (Pareto front).

[0094] Typical optimized solutions: Solution A: 60 micrometers wavy spacing, 8 micrometers thinning, 18 micrometers total copper thickness (85% yield, 45 minutes production time); Solution B: 50 micrometers wavy spacing, 10 micrometers thinning, 15 micrometers total copper thickness (88% yield, 50 minutes production time).

[0095] The optimal solution selection is based on production scheduling priorities (such as prioritizing efficiency for urgent orders and prioritizing yield for high value-added products), automatically selecting the optimal solution from the Pareto frontier, and supporting process engineers to visually adjust and confirm the selection.

[0096] In some embodiments, in order to address the problem that traditional film design relies on finite element simulation, but it is difficult to cover all substrate variations (such as surface roughness and differences in copper layer hardness), resulting in a decrease in the bonding effect of the "standard design" in personalized scenarios (such as high-roughness substrates).

[0097] By constructing a GAN-driven film edge design system, and by learning from historical excellent film designs (low wrinkles, high adhesion rate), and combining the real-time physical parameters of the current substrate (surface roughness, break height, copper layer hardness), an adaptive edge shape (personalized parameters of inverted wavy / serrated) is generated, achieving dynamic matching of "substrate-design".

[0098] The GAN model construction includes: Generator: The input layer receives three substrate parameters (surface roughness Ra, discontinuity height H, and copper layer hardness HV), and generates a geometric parameter matrix of the film edge (wave spacing, wave depth, and sawtooth angle, resolution 10×10 pixels, corresponding to an actual size of 50-100 micrometers) through a 5-layer convolutional neural network. Discriminator: The input layer receives the film edge design drawing and corresponding bonding effect data (wrinkle rate, number of bubbles), and outputs the probability (0-1) of whether the design is excellent through a 3-layer fully connected network. Loss function: "Adversarial loss + bonding effect loss" is used. The adversarial loss optimizes the realism of the generator, while the bonding effect loss (such as the MSE of the wrinkle rate) guides the generator towards optimization towards "low defects".

[0099] Data training involves collecting multiple sets of historically excellent film designs (wrinkle rate ≤5%, number of bubbles ≤2 / ㎡), corresponding substrate parameters (Ra: 0.5-2.0μm, H: 5-15μm, HV: 120-180), and bonding effect data to construct a training set.

[0100] The GAN model was trained using the Adam optimizer for 5000 iterations until the edge design generated by the generator had an “excellent probability” of ≥90% in the discriminator and the fitting effect loss was ≤0.01.

[0101] The adaptive design process includes: During production, three parameters of the current substrate are collected in real time using a laser roughness meter (Ra accuracy ±0.1μm), a metallographic microscope (H accuracy ±1μm), and a Vickers hardness tester (HV accuracy ±5), and input into the generator. The generator outputs personalized edge parameters (e.g., when Ra=1.5μm and H=12μm, it generates a wavy pattern spacing of 70μm, a wave depth of 35μm, and a sawtooth angle of 100°), which are automatically imported into a laser plotter to create film. After the film is created, the stress distribution is quickly verified through finite element simulation (time ≤1 minute). If the stress concentration factor is ≤1.5, production begins; otherwise, the generator readjusts the parameters (increasing the wave density by 20%) until the requirements are met.

[0102] In some embodiments, to address the problem that traditional dry film lamination relies on "post-launch inspection" (checking for wrinkles / bubbles after lamination), which cannot detect dynamic stress changes during the lamination process in real time, resulting in a high scrap rate due to the inability to intervene in time after defects occur, a digital twin of dry film lamination is constructed. This digital twin integrates physical models (mechanical response of dry film-copper layer-substrate) with real-time data (lamination pressure, temperature, substrate deformation) to simulate the lamination process in real time, provide early warning of defects, and automatically adjust process parameters.

[0103] The construction of the digital twin includes: using finite element software such as ANSYS to build a three-dimensional finite element model of the dry film (polyimide), copper layer (electrolytic copper), and substrate (FR4), defining the material constitutive relationship (the elastic modulus of the dry film changes with temperature: 2.5GPa at 120℃ and 1.8GPa at 150℃), and simulating stress transfer and deformation during pressing.

[0104] The data model uses the MES system to collect real-time data on the pressure (accuracy ±0.1MPa), temperature (accuracy ±0.5℃), conveyor speed (accuracy ±0.1m / min) of the pressing equipment, as well as the real-time deformation data of the substrate (collected using a laser displacement sensor, accuracy ±0.01mm).

[0105] The fusion model uses a digital twin platform built with Python to synchronize the physical model with real-time data (latency ≤100ms) and generate a dynamic stress cloud map of the pressing process (resolution 100×100 pixels).

[0106] Real-time early warning and intervention are achieved by setting stress thresholds (maximum stress of dry film ≤ 3MPa, stress at the interface between copper layer and substrate ≤ 2MPa). When the stress simulated by the digital twin exceeds the threshold, the system automatically triggers the following: Parameter adjustment: reducing the pressing pressure (10%), increasing the temperature (5℃), and slowing down the conveyor speed (10%), adjusting equipment parameters in real time through a PID algorithm; Visual early warning: marking stress concentration areas on the monitoring interface (e.g., "stress exceeds standard in the 100×100mm area in the upper left corner") to prompt the operator to check carefully; Data recording: saving the current parameters and stress cloud map for subsequent process optimization.

[0107] After each batch of production is completed, the actual defect data (wrinkle location, number of bubbles) is compared with the digital twin simulation results. Bayesian optimization is used to adjust the material parameters of the physical model (such as the elastic modulus correction coefficient of the dry film) to reduce simulation errors and improve prediction accuracy.

[0108] In some embodiments, in order to address the problem that traditional etching parameters (concentration, time, pressure) rely on "experience-based settings" and cannot cope with variations in dry film adhesion (such as changes in etching resistance caused by local wrinkles), resulting in large fluctuations in lateral etching amount (5-12μm) and high etching incompleteness rate in the two-electric region (approximately 8%).

[0109] By constructing a reinforcement learning agent, the agent receives real-time data on dry film bonding status (wrinkle rate, number of bubbles) and etching process (etching solution concentration, temperature, copper layer thickness), and dynamically adjusts etching parameters to minimize lateral etching (target ≤5μm) and maximize etching cleanliness (target ≥99%) as optimization objectives.

[0110] The reinforcement learning model design includes: State space: containing 6 dimensions: dry film wrinkle rate (%), number of bubbles (number / m²) 2 Etching solution concentration (%), etching solution temperature (°C), current copper layer thickness (μm), and spray pressure (MPa).

[0111] Action Space: Includes 3 discrete actions: Etching solution concentration: +1%, 0, -1% (range 4-8%); Etching time: +10s, 0, -10s (range 60-200s); Spray pressure: +0.1MPa, 0, -0.1MPa (range 0.3-0.7MPa).

[0112] Reward function: Reward=100 5×(lateral corrosion amount) 5) 10×(etching incomplete rate) + 2×(concentration saving rate) (5 points are penalized for every 1μm increase in side etching amount beyond the target; 10 points are penalized for every 1% decrease in etching incomplete rate; 2 points are awarded for every 1% decrease in concentration, balancing yield and cost).

[0113] Model training involves collecting multiple sets of etching process data (including states, actions, and rewards) and training the agent using the proximal policy optimization (PPO) algorithm (PPO is suitable for continuous action spaces and has better stability than DQN). During training, Monte Carlo simulations are used to generate virtual etching scenarios (such as extreme cases with "10% wrinkling rate and 5% concentration") to improve the agent's generalization ability. Training continues until the reward function converges (reward fluctuation ≤5% over 1000 consecutive iterations). At this point, the agent's prediction error for lateral etching amount is ≤1μm, and its prediction error for etching incompleteness rate is ≤1%.

[0114] The real-time control process includes: Before etching, the dry film bonding status (wrinkle rate, number of bubbles) is acquired through a machine vision system (multimodal detection in Example 5) and input into the agent. During etching, the online copper thickness detection sensor (X-ray thickness gauge) collects the copper layer thickness every 5 seconds, and the etching solution concentration sensor (refractometer) monitors the concentration in real time. All data is synchronized to the agent. The agent outputs the optimal action based on the current status (e.g., "concentration +1%, time +10s, pressure +0.1MPa"), and adjusts the etching equipment parameters through the PLC controller. After etching, the lateral etching amount and etching incompleteness rate are detected by a metallographic microscope, and the results are fed back to the agent to update the policy (the model weights are adjusted every 10 batches).

[0115] In some embodiments, in order to address the problem that traditional dry film bonding inspection relies on a single vision (taking a surface image), which cannot identify latent defects (such as "virtual bonding": the surface has no wrinkles but the internal air bubbles are not completely expelled), causing "latent defects" to be transformed into "obvious defects" (such as side etching, incomplete etching) during subsequent etching.

[0116] By constructing a multimodal fusion detection system, which integrates data from three modalities—machine vision (surface image), infrared thermal imaging (temperature distribution), and pressure sensor (pressure distribution)—and extracting cross-modal features through the Transformer model, the system accurately identifies "obvious defects" (wrinkles, bubbles) and "latent defects" (false bonding).

[0117] Modal data acquisition includes: Machine vision: A linear CCD camera (2000 dpi resolution) is installed at the outlet of the laminating equipment to capture images of the dry film surface (each frame covering a 100mm × 100mm area) for detecting wrinkles (discontinuous edges) and bubbles (circular bright areas). Infrared thermal imaging: An uncooled infrared thermal imager (320 × 240 resolution, temperature accuracy ±0.5℃) is installed to capture the temperature distribution on the dry film surface. "Virtually bonded" areas have a temperature 2-5℃ higher than normal areas due to poor adhesion and slow heat conduction. Pressure sensors: An array of pressure sensors (10 × 10 array, accuracy ±0.01 MPa) is installed under the laminating platform to collect the pressure distribution during lamination. "Virtually bonded" areas have a pressure 10-20% lower than normal areas because the dry film is not fully compacted.

[0118] The multimodal fusion model uses a Transformer Encoder as the fusion backbone to encode features for each modality: visual modality: image features are extracted using ResNet-50 (output 2048-dimensional vector); infrared modality: temperature distribution features are extracted using a CNN (output 1024-dimensional vector); pressure modality: pressure distribution features are extracted using an MLP (output 512-dimensional vector). The feature vectors of the three modalities are concatenated (2048+1024+512=3584 dimensions) and input into the Transformer Encoder (6 layers, 8-head self-attention) to learn cross-modal associations (e.g., "high temperature + low pressure" corresponds to "virtual patch"). The output layer uses softmax classification to output the probability of four types of defects: wrinkles (0-1), bubbles (0-1), virtual patches (0-1), and normal (0-1).

[0119] Real-time detection and intervention: The system processes multimodal data every 2 seconds (delay ≤ 1 second). When the probability of defects in a certain area exceeds the threshold (wrinkles ≥ 0.7, bubbles ≥ 0.6, false bonding ≥ 0.5), the following is triggered: Local marking: Defect areas are marked with different colors on the monitoring interface (wrinkles red, bubbles yellow, false bonding blue); Automatic rework: The conveyor is controlled to transport the substrate to the secondary pressing station, and local high pressure (15% higher than the standard pressure) is applied to the defect area, maintaining 130°C for 10 seconds to eliminate false bonding and bubbles; Data recording: The multimodal data and processing results of the defect area are saved for subsequent model optimization (Transformer weights are updated every 1000 sets of data).

[0120] Please see Figure 2 As shown, Figure 2 This is a schematic diagram of a system 200 for improving the dry film bonding effect and etching quality in adhesive-free delaminated areas of a PCB, as provided in this application embodiment. The system 200 is used to perform the steps of the methods for improving the dry film bonding effect and etching quality in adhesive-free delaminated areas of a PCB as described in the above embodiments. The system 200 can be a single server or a server cluster, or it can be a terminal, such as a handheld terminal, laptop computer, wearable device, or robot.

[0121] like Figure 2 As shown, the system 200 for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB includes: Film design unit 201 is used to design the film used for pattern electroplating. It changes the straight line design of the edge of the non-adhesive area to an inverted wave pattern or serrated design to disperse and release the stress generated during dry film lamination, and improve the coverage and adhesion tightness of the dry film in the step area. The total amount reduction unit 202 is used to uniformly reduce the thickness of the bottom copper after the outer substrate is laminated and before the dry film is pressed, thereby reducing the total amount of copper that needs to be etched away and reducing the risk of side etching caused by the dry film not being pressed firmly. The thickness reduction unit 203 is used to reduce the total thickness of the bottom copper and the top copper of the substrate to 15-20 micrometers, reduce the total copper height after pattern electroplating, reduce the absolute difference height between the first and second electroplating, and make the dry film easy to press tightly and without bubbles.

[0122] In some embodiments, the film used for the design pattern electroplating changes the straight line design of the edge of the adhesive-free area to an inverted wavy or sawtooth design to disperse and release the stress generated during dry film lamination. This includes: simulating the stress distribution during the dry film lamination process using a finite element analysis algorithm, and generating an inverted wavy or sawtooth profile of the edge of the adhesive-free area based on the simulation results; wherein the wave crest spacing, wave depth, or sawtooth angle of the wavy pattern is automatically calculated and determined according to the substrate thickness and the copper layer discontinuity height to maximize stress dispersion.

[0123] In some embodiments, the process of uniformly thinning the bottom copper thickness after laminating the outer substrate and before drying the film includes: acquiring copper layer thickness data on the substrate surface using an online copper thickness detection sensor, calculating the micro-etching time or copper stripping depth of the copper reduction process using a closed-loop control algorithm, and establishing a mapping relationship between the amount of bottom copper thickness reduction and the difficulty of subsequent etching.

[0124] In some embodiments, reducing the total thickness of the base copper and surface copper of the substrate to 15-20 micrometers and reducing the total copper height after pattern electroplating includes: establishing a mathematical model of the total copper thickness and the difference height based on historical data of the first and second electroplating breaks using a regression analysis algorithm; adjusting the thickness ratio of the base copper and surface copper according to the optimal solution output by the model, so that the total thickness is controlled within the range of 15-20 micrometers, while meeting the conductivity requirements.

[0125] In some embodiments, the method further includes: during the film design stage, using machine learning algorithms to train a dry film wrinkle prediction model, inputting parameters such as the shape of the glue-free area and the copper layer discontinuity height, and outputting the optimal structural parameters for the inverted wave pattern or sawtooth design, so as to predict and reduce stress concentration during pressing in advance.

[0126] In some embodiments, the method further includes: in the copper reduction process, acquiring images of the substrate surface in real time through a machine vision system, using an image recognition algorithm to detect the uniformity of the copper layer after thinning, and if a local thickness abnormality is detected, automatically adjusting the spray pressure of the micro-etching equipment or the travel speed of the copper planing tool to achieve dynamic compensation.

[0127] In some embodiments, the method further includes: establishing a process database of dry film lamination effect and copper thickness structure, analyzing historical yield data through data mining algorithms, generating a copper thickness reduction strategy library for different break heights, and automatically matching the optimal combination of bottom copper and surface copper thicknesses during new batch production.

[0128] In some embodiments, the method further includes: after pressing the dry film, using infrared thermal imaging technology to detect the dry film bonding temperature distribution in the step area, calculating the stress value of the temperature abnormal area through a heat conduction model algorithm, and automatically adjusting the temperature or pressure parameters of the pressing equipment for secondary pressing if the stress value exceeds a preset threshold.

[0129] In some embodiments, the method further includes: predicting the probability of lateral etching caused by poor dry film pressing using a neural network algorithm before the etching process, and dynamically adjusting the etching solution concentration or etching time based on the prediction results to form intelligent linkage control of pressing effect and etching parameters.

[0130] In some embodiments, the method further includes: using a multi-objective optimization algorithm to collaboratively optimize the film pattern design parameters, copper thickness reduction, and total copper thickness, with dry film bonding tightness, etching yield, and production efficiency as optimization objectives, to generate a globally optimal combination of process parameters.

[0131] It should be noted that those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the system and its modules for improving the dry film bonding effect and etching quality in the glue-free delamination area of ​​a PCB, as described above, can be found in the corresponding contents of the various embodiments of the method for improving the dry film bonding effect and etching quality in the glue-free delamination area of ​​a PCB, and will not be repeated here.

[0132] The aforementioned methods for improving the dry film lamination effect and etching quality in glue-free delamination areas of PCBs can be implemented as a computer program, which can be used in, for example... Figure 2 It runs on the device shown.

[0133] Please see Figure 3 , Figure 3 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application. The computer device includes a processor, a memory, and a network interface connected via a device bus, wherein the memory may include a storage medium and internal memory.

[0134] The storage medium may store operating devices and computer programs. The computer program includes program instructions that, when executed, cause the processor to perform any method to improve the dry film bonding effect and etching quality in the glue-free delamination areas of the PCB.

[0135] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0136] The internal memory provides an environment for the execution of computer programs in non-volatile storage media. When the computer program is executed by the processor, it enables the processor to perform any method to improve the dry film bonding effect and etching quality in the glue-free delamination area of ​​the PCB.

[0137] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 3The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0139] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps: The film used for graphic electroplating was designed by changing the straight lines at the edges of the non-adhesive areas to inverted wavy or sawtooth patterns to disperse and release the stress generated during dry film lamination, thereby improving the coverage and adhesion tightness of the dry film in the stepped areas. After the outer substrate is laminated and before the dry film is pressed, the thickness of the bottom copper is uniformly reduced to reduce the total amount of copper that needs to be etched away later, thereby reducing the risk of side etching due to insufficient dry film pressing. The total thickness of the base copper and surface copper of the substrate is reduced to 15-20 micrometers, which reduces the total copper height after pattern electroplating and reduces the absolute height difference between the first and second electroplating processes, making it easier to press the dry film tightly and without bubbles.

[0140] In some embodiments, the film used for the design pattern electroplating changes the straight line design of the edge of the adhesive-free area to an inverted wavy or sawtooth design to disperse and release the stress generated during dry film lamination. This includes: simulating the stress distribution during the dry film lamination process using a finite element analysis algorithm, and generating an inverted wavy or sawtooth profile of the edge of the adhesive-free area based on the simulation results; wherein the wave crest spacing, wave depth, or sawtooth angle of the wavy pattern is automatically calculated and determined according to the substrate thickness and the copper layer discontinuity height to maximize stress dispersion.

[0141] In some embodiments, the process of uniformly thinning the bottom copper thickness after laminating the outer substrate and before drying the film includes: acquiring copper layer thickness data on the substrate surface using an online copper thickness detection sensor, calculating the micro-etching time or copper stripping depth of the copper reduction process using a closed-loop control algorithm, and establishing a mapping relationship between the amount of bottom copper thickness reduction and the difficulty of subsequent etching.

[0142] In some embodiments, reducing the total thickness of the base copper and surface copper of the substrate to 15-20 micrometers and reducing the total copper height after pattern electroplating includes: establishing a mathematical model of the total copper thickness and the difference height based on historical data of the first and second electroplating breaks using a regression analysis algorithm; adjusting the thickness ratio of the base copper and surface copper according to the optimal solution output by the model, so that the total thickness is controlled within the range of 15-20 micrometers, while meeting the conductivity requirements.

[0143] In some embodiments, the method further includes: during the film design stage, using machine learning algorithms to train a dry film wrinkle prediction model, inputting parameters such as the shape of the glue-free area and the copper layer discontinuity height, and outputting the optimal structural parameters for the inverted wave pattern or sawtooth design, so as to predict and reduce stress concentration during pressing in advance.

[0144] In some embodiments, the method further includes: in the copper reduction process, acquiring images of the substrate surface in real time through a machine vision system, using an image recognition algorithm to detect the uniformity of the copper layer after thinning, and if a local thickness abnormality is detected, automatically adjusting the spray pressure of the micro-etching equipment or the travel speed of the copper planing tool to achieve dynamic compensation.

[0145] In some embodiments, the method further includes: establishing a process database of dry film lamination effect and copper thickness structure, analyzing historical yield data through data mining algorithms, generating a copper thickness reduction strategy library for different break heights, and automatically matching the optimal combination of bottom copper and surface copper thicknesses during new batch production.

[0146] In some embodiments, the method further includes: after pressing the dry film, using infrared thermal imaging technology to detect the dry film bonding temperature distribution in the step area, calculating the stress value of the temperature abnormal area through a heat conduction model algorithm, and automatically adjusting the temperature or pressure parameters of the pressing equipment for secondary pressing if the stress value exceeds a preset threshold.

[0147] In some embodiments, the method further includes: predicting the probability of lateral etching caused by poor dry film pressing using a neural network algorithm before the etching process, and dynamically adjusting the etching solution concentration or etching time based on the prediction results to form intelligent linkage control of pressing effect and etching parameters.

[0148] In some embodiments, the method further includes: using a multi-objective optimization algorithm to collaboratively optimize the film pattern design parameters, copper thickness reduction, and total copper thickness, with dry film bonding tightness, etching yield, and production efficiency as optimization objectives, to generate a globally optimal combination of process parameters.

[0149] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the method for improving the dry film bonding effect and etching quality in glue-free delamination areas of a PCB, as provided in any embodiment of this application.

[0150] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0151] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for improving the dry film lamination effect and etching quality in glue-free delamination areas of a PCB, characterized in that, include: The film used for graphic electroplating was designed by changing the straight lines at the edges of the non-adhesive areas to inverted wavy or sawtooth patterns to disperse and release the stress generated during dry film lamination, thereby improving the coverage and adhesion tightness of the dry film in the stepped areas. After the outer substrate is laminated and before the dry film is pressed, the thickness of the bottom copper is uniformly reduced to reduce the total amount of copper that needs to be etched away later, thereby reducing the risk of side etching due to insufficient dry film pressing. The total thickness of the base copper and surface copper of the substrate is reduced to 15-20 micrometers, which reduces the total copper height after pattern electroplating and reduces the absolute height difference between the first and second electroplating processes, making it easier to press the dry film tightly and without bubbles.

2. The method according to claim 1, characterized in that, The film used for electroplating the designed pattern will have its straight lines at the edges of the adhesive-free areas changed to a wavy or sawtooth pattern to disperse and release the stress generated during dry film lamination, including: The stress distribution during the dry film lamination process is simulated using finite element analysis algorithms. Based on the simulation results, inverted wavy or sawtooth contours are generated at the edges of the glue-free areas. The wave crest spacing, wave depth, or sawtooth angle of the wavy pattern are automatically calculated and determined according to the substrate thickness and the copper layer discontinuity height to maximize stress dispersion.

3. The method according to claim 1, characterized in that, The process of uniformly thinning the bottom copper layer after laminating the outer substrate and before drying the film includes: Online copper thickness detection sensors are used to obtain copper layer thickness data on the substrate surface. A closed-loop control algorithm is used to calculate the micro-etching time or copper removal depth of the copper reduction process, so that the amount of copper thickness reduction is mapped to the difficulty of subsequent etching.

4. The method according to claim 1, characterized in that, The method of reducing the total thickness of the base copper and surface copper of the substrate to 15-20 micrometers, thereby reducing the total copper height after pattern electroplating, includes: Based on historical data of the difference height between the first and second electrical circuits, a mathematical model of the total copper thickness and the difference height was established through regression analysis algorithm. The thickness ratio of the bottom copper and the top copper was adjusted according to the optimal solution output by the model, so that the total thickness was controlled within the range of 15-20 micrometers, while meeting the conductivity requirements.

5. The method according to claim 1, characterized in that, The method further includes: During the film design phase, a dry film wrinkle prediction model is trained using machine learning algorithms. Parameters such as the shape of the glue-free area and the height of the copper layer break are input, and the optimal structural parameters for the inverted wave pattern or sawtooth design are output to predict and reduce stress concentration during lamination in advance.

6. The method according to claim 1, characterized in that, The method further includes: In the copper reduction process, the machine vision system acquires images of the substrate surface in real time and uses image recognition algorithms to detect the uniformity of the copper layer after thinning. If local thickness abnormalities are detected, the spray pressure of the micro-etching equipment or the travel speed of the copper planing tool is automatically adjusted to achieve dynamic compensation.

7. The method according to claim 1, characterized in that, The method further includes: A process database of dry film lamination effect and copper thickness structure is established. Historical yield data is analyzed through data mining algorithms to generate a copper thickness reduction strategy library for different break heights. The optimal combination of bottom copper and surface copper thickness is automatically matched during new batch production.

8. The method according to claim 1, characterized in that, The method further includes: After the dry film is pressed, infrared thermal imaging technology is used to detect the temperature distribution of the dry film bonding in the step area. The stress value of the temperature abnormal area is calculated by the heat conduction model algorithm. If it exceeds the preset threshold, the temperature or pressure parameters of the pressing equipment are automatically adjusted for secondary pressing.

9. The method according to claim 1, characterized in that, The method further includes: Before the etching process, the probability of side etching caused by poor dry film pressing is predicted by a neural network algorithm. The concentration of the etching solution or the etching time is dynamically adjusted based on the prediction results, forming an intelligent linkage control between the pressing effect and the etching parameters.

10. The method according to claim 1, characterized in that, The method further includes: A multi-objective optimization algorithm is used to collaboratively optimize the film pattern design parameters, copper thickness reduction, and total copper thickness. The optimization objectives are dry film bonding tightness, etching yield, and production efficiency, generating the globally optimal combination of process parameters.

Images