Window-free mark point structure for PCB cover film and manufacturing method thereof

Abstract

The application relates to the field of integrated circuit manufacturing, in particular to the technical field of manufacturing special-purpose equipment such as special-purpose photoetching machines and etching machines for semiconductor devices, and provides a window-free Mark point structure for a PCB cover film and a manufacturing method thereof. The method comprises the following steps: preparing a PCB substrate and pretreating a copper surface; for a yellow cover film area of the PCB substrate, the yellow cover film is attached to a corresponding position of the PCB substrate, the cover film at a Mark point position is not subjected to windowing treatment, and the color of the yellow cover film itself is used to form contrast with a background copper surface for SMT equipment identification; for a black cover film area, a white ink point is first manufactured at a Mark point position of the substrate, and the black cover film is attached to a corresponding position of the PCB substrate; and after the attachment is completed, the cover film is subjected to solidification treatment. Through the window-free design of the yellow cover film area, the drilling / laser cutting process and the nickel gold sinking process are omitted, and the consumption of precious metals is avoided.

Description

Technical Field

[0001] This application relates to the field of integrated circuit manufacturing, and particularly to the field of manufacturing technology for special equipment for semiconductor devices such as dedicated lithography machines and etching machines. It relates to a windowless Mark point structure for PCB cover film and its fabrication method. Background Technology

[0002] In the field of integrated circuit manufacturing, especially in the production of specialized semiconductor device equipment such as lithography machines and etching machines, the traditional method of creating cover film mark points typically involves opening windows in the cover film and performing an immersion nickel-gold plating process on the exposed copper surface to form mark points that can be identified by SMT equipment. This process has the following significant drawbacks:

[0003] 1. High cost: Opening windows requires drilling or laser cutting processes, which increases equipment wear and processing time; the immersion nickel-gold process consumes a large amount of gold salt, resulting in high material costs.

[0004] 2. Complex process: After opening the window, multiple additional processes such as nickel plating and surface treatment are required, which is complicated and prone to misalignment.

[0005] 3. Marking is easily damaged: During the pretreatment of the grinding plate, the Mark points in the windowed area are prone to deformation due to the loss of the protective film; the traditional FQC blackening process has problems with unstable coloring and easy diffusion and fading, which affects the reliability of recognition.

[0006] 4. Poor process compatibility: When the black cover film area adopts the windowed immersion gold design, the reflective effect of the gold surface is easily affected by the surface condition, resulting in SMT equipment recognition errors; the punching process requires the design of special punches for the windowed structure, increasing mold costs.

[0007] Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the Invention

[0008] This application provides a windowless Mark dot structure for PCB cover film and its manufacturing method, aiming to solve the problems in the prior art. The improvement ideas for the above-mentioned problems mainly focus on optimizing the immersion gold process or adjusting the windowing accuracy, but none of them break through the technical framework of "windowing immersion gold is necessary". For example, it does not propose to use the contrast between the color of the cover film itself and the copper surface to replace gold dot identification, nor does it address the issue of marking by the color contrast between white ink dots and black cover film.

[0009] In a first aspect, embodiments of this application provide a method for fabricating a windowless Mark dot structure for a PCB cover film, the method comprising:

[0010] Prepare the PCB substrate and perform pretreatment on the copper surface of the PCB substrate.

[0011] For the yellow cover film area of ​​the PCB substrate, the yellow cover film is attached to the corresponding position of the PCB substrate. The cover film at the Mark point position is not opened. The color of the yellow cover film itself is used to form a contrast with the background copper surface for SMT equipment to identify. For the black cover film area, white ink dots are first made at the Mark point position of the PCB substrate. The black cover film is then attached to the corresponding position of the PCB substrate to expose the white ink dots.

[0012] After lamination, the yellow and black cover films are cured. Multiple sensors deployed on the production line collect real-time process parameter data on temperature, pressure, and position. This data is analyzed in real time, and when deviations from preset ranges are detected, an early warning signal is issued, and corresponding production equipment parameters are adjusted to ensure the quality of the Mark point structure and the stability of the production process. Before creating the Mark points, a machine learning-based intelligent algorithm is used to analyze and train the contrast between the cover film and the background copper surface for different color combinations and the SMT equipment recognition success rate in historical production data. A contrast-recognition success rate prediction model is established. Based on this model, it is predicted whether the contrast between the yellow cover film and the background copper surface, and between the white ink dots and the black cover film in the current design meets the recognition requirements of the SMT equipment. The cover film color parameters or Mark point design parameters are automatically adjusted based on the prediction results.

[0013] In some embodiments, the preparation of the PCB substrate and the pretreatment of the copper surface of the PCB substrate include: grinding the copper surface of the PCB substrate with a computer-controlled mechanical grinding device to remove the oxide layer and impurities on the copper surface, or degreasing and micro-etching the copper surface with a chemical cleaning line to make the copper surface reach a surface state suitable for subsequent lamination of the cover film.

[0014] In some embodiments, attaching the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area includes: using a computer-controlled vacuum pressing device to precisely attach the yellow cover film to the designated area of ​​the PCB substrate according to preset temperature, pressure and time parameters, and ensuring the accurate position of the cover film through the alignment system of the device during the pressing process.

[0015] In some embodiments, the cover film at the Mark point position is not subjected to windowing processing, including: during the design and manufacturing process of the cover film, the Mark point position is marked by computer-aided design software, processing data of the cover film is generated, the processing equipment is controlled not to perform drilling or laser cutting windowing operations at the Mark point position, and at the same time, a vision alignment system is used to ensure that the cover film at the Mark point position is accurately aligned with the target position on the PCB substrate.

[0016] In some embodiments, the step of using the color of the yellow cover film itself to create a contrast with the background copper surface for SMT equipment recognition includes: before the cover film is applied, acquiring color images of the yellow cover film and the background copper surface through a computer image acquisition system, analyzing the contrast value of the two using an image processing algorithm, and confirming that the design meets the requirements when the contrast value reaches a preset threshold that the SMT equipment can recognize; otherwise, adjusting the cover film color or the copper surface processing technology.

[0017] In some embodiments, for the black cover film area, firstly, white ink dots are made at the Mark point positions on the PCB substrate, including: using a computer-controlled inkjet printing device, according to the preset Mark point coordinates and size data, precisely spraying white ink at the Mark point positions on the PCB substrate to form white ink dots, and during the printing process, the position and shape of the ink dots are monitored in real time by the device's vision inspection system to ensure that they meet the design requirements.

[0018] In some embodiments, the step of attaching the black cover film to the corresponding position on the PCB substrate to expose the white ink dots includes: using a computer-controlled bonding device, identifying the position of the white ink dots on the PCB substrate using a vision alignment system, accurately aligning the pre-opened clearance window on the black cover film corresponding to the white ink dots with the white ink dots, and then bonding them to ensure that the white ink dots are fully exposed and the cover film is flat.

[0019] In some embodiments, after the bonding is completed, the yellow cover film and the black cover film are cured, including: placing the bonded PCB substrate into a computer-controlled curing oven, and performing heating, heat preservation and cooling treatments according to a preset temperature-time curve. During the curing process, the temperature data is monitored in real time by a temperature sensor in the oven, and the heating power is automatically adjusted by the control system to ensure that the cover film is completely cured and has stable performance.

[0020] In some embodiments, the step of using machine learning-based intelligent algorithms to analyze and train the contrast of cover film and background copper surface with different color combinations and the SMT equipment recognition success rate in historical production data to establish a contrast-recognition success rate prediction model includes: acquiring cover film color parameters, background copper surface processing parameters, actual measured contrast values, and corresponding SMT equipment recognition success rate data stored in the historical production database through a computer data management system; cleaning and normalizing the acquired data using a computer preprocessing module to remove outlier data points and unify the data format; constructing a training dataset by using a computer feature engineering module with cover film color parameters, background copper surface processing parameters, and contrast values ​​as input features and SMT equipment recognition success rate as the output label; selecting a regression model or classification model from a preset machine learning algorithm library using a computer algorithm selection module, and optimizing the model parameters of the regression model or classification model through cross-validation; and iteratively training the selected regression model or classification model using a computer training module until the prediction error of the regression model or classification model on the test set is lower than a preset threshold, thereby obtaining the contrast-recognition success rate prediction model.

[0021] Secondly, this application provides a windowless Mark dot structure for PCB cover film, which is manufactured based on the method provided in any embodiment of this application.

[0022] 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

[0023] 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.

[0024] Figure 1 This is a schematic flowchart illustrating the steps of a method for fabricating a windowless Mark dot structure for a PCB cover film according to an embodiment of this application;

[0025] Figure 2 This is a schematic block diagram of a fabrication system for a windowless Mark dot structure for PCB cover film provided in one embodiment of this application;

[0026] Figure 3 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0027] 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

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0034] In the field of printed circuit board (PCB) manufacturing, the traditional method of creating cover film markers typically involves creating windows in the cover film and applying an immersion nickel-gold plating to the exposed copper surface to form markers that can be identified by surface mount technology (SMT) equipment. This process has the following significant drawbacks:

[0035] High cost: Opening windows requires drilling or laser cutting processes, which increases equipment wear and processing time; the immersion nickel-gold process consumes a large amount of gold salt, keeping material costs high.

[0036] The process is complex: after opening the window, multiple additional processes such as nickel plating and surface treatment are required, which is cumbersome and prone to misalignment.

[0037] Marking is easily damaged: During the pretreatment of the grinding plate, the Mark points in the windowed area are prone to deformation due to the loss of the protective film; the traditional FQC blackening process has problems such as unstable coloring and easy diffusion and fading, which affects the reliability of recognition.

[0038] Poor process compatibility: When the black cover film area adopts the windowed immersion gold design, the reflective effect of the gold surface is easily affected by the surface condition, resulting in SMT equipment recognition errors; the punching process requires the design of special punches for the windowed structure, increasing mold costs.

[0039] In existing technologies, improvements to address the aforementioned issues primarily focus on optimizing the immersion gold process or adjusting the precision of the windowing process, but none of these approaches break through the technical framework of "windowing is mandatory for immersion gold." For example, no solution has been proposed that utilize the contrast between the cover film's own color and the copper surface to replace gold dot identification, nor has a scheme been developed to achieve marking through the color contrast between white ink dots and the black cover film. Furthermore, existing processes lack quantitative analysis of the success rate of mark dot recognition and color contrast, and have not incorporated intelligent algorithms into the process design and monitoring workflow.

[0040] To solve the above problem, please refer to Figure 1 This application provides a method for fabricating a windowless Mark dot structure for PCB cover film, applicable to computer equipment, which can be deployed on a single server or server cluster. It can also be deployed on handheld terminals, laptops, wearable devices, or robots, etc.

[0041] The provided method for fabricating a windowless Mark dot structure for PCB cover film includes steps S101 to S103. Details are as follows:

[0042] Step S101. Prepare the PCB substrate and perform pretreatment on the copper surface of the PCB substrate.

[0043] Specifically, the purpose of this step is to remove contaminants (such as oxides, grease, dust, etc.) from the copper surface of the PCB substrate, and to enhance the adhesion between the copper surface and the cover film by roughening the surface, thereby ensuring the reliability of the subsequent bonding process.

[0044] The cleaning process removes surface impurities through chemical cleaning or physical grinding, avoiding any impact on the adhesion between the cover film and the copper surface.

[0045] Roughening treatment increases the roughness of the copper surface through micro-etching or sandblasting processes, forming a micro-anchor point structure and improving mechanical adhesion.

[0046] Chemical cleaning involves immersing the PCB substrate in an acidic or alkaline cleaning solution (such as a sulfuric acid-hydrogen peroxide mixture or sodium hydroxide solution) to remove oxides and organic contaminants through a chemical reaction. Typical parameters: temperature 40-60℃, immersion time 2-5 minutes, enhanced cleaning effect by combining with ultrasonic vibration.

[0047] Physical grinding uses a grinding machine equipped with a nylon brush or volcanic ash abrasive to mechanically grind the copper surface, with the roughness controlled at Ra 1.0-2.0μm (detected by a surface roughness meter).

[0048] After washing, rinse thoroughly with deionized water to remove residual chemicals and avoid ion contamination; then dry with hot air (80-100℃) or vacuum to ensure the surface is dry and free of moisture.

[0049] Step S102. For the yellow cover film area of ​​the PCB substrate, the yellow cover film is attached to the corresponding position of the PCB substrate. The cover film at the Mark point position is not opened. The color of the yellow cover film itself is used to form a contrast with the background copper surface for SMT equipment to identify. For the black cover film area, white ink dots are first made at the Mark point position of the PCB substrate. The black cover film is attached to the corresponding position of the PCB substrate so that the white ink dots are exposed.

[0050] Specifically, the yellow cover film area uses a windowless design to directly utilize the color contrast (yellow and copper) between the yellow cover film and the copper surface to form a mark point for SMT equipment to identify.

[0051] The black overlay area replaces the traditional windowed gold-plated Mark point with a "white oil point + black overlay" structure, using the high contrast (white and black) between the white oil point (white ink) and the black overlay to replace the gold point's reflective identification.

[0052] The yellow cover film area should use a yellow cover film with low light transmittance and high color density (such as PI material, 25-50μm thick). Its color number needs to be pre-verified by contrast (e.g., the gray difference ΔE with the copper surface is ≥15, which is detected by a spectrophotometer).

[0053] Using an automated bonding machine (such as a COG bonding machine), a vision alignment system (CCD camera) identifies the pre-positioning marks on the substrate's Mark points. The yellow cover film is then aligned with the corresponding area on the PCB substrate, ensuring that the Mark point location has no openings and completely covers the copper surface. The bonding pressure is controlled at 0.1-0.3 MPa, and the temperature at 50-80℃ (adjusted according to the adhesive properties of the cover film) to prevent stretching and deformation of the cover film.

[0054] For the black overlay area, high-contrast white ink (such as curable and chemically resistant epoxy resin white ink) is used to print circular or square white ink dots at the Mark points on the PCB substrate using a precision screen printing stencil (200-300T mesh). The size is typically φ0.5-1.0mm (consistent with traditional gold dots). After printing, pre-baking (70-90℃, 15-20 minutes) and full curing (150-180℃, 30-60 minutes) are performed to ensure that the adhesion of the white ink dots is ≥3M tape method 5B standard (GB / T 9286-1998).

[0055] Black cover film is used for bonding (e.g., PI material with added carbon black, optical density ≥3.0). An automated bonding machine aligns the cover film with the substrate, fully exposing the white paint dot area (i.e., the cover film does not need to be windowed at the white paint dot location, directly covering the non-mark area). The bonding process parameters are the same as for the yellow cover film, ensuring an alignment accuracy of ≤±50μm between the white paint dot edge and the cover film.

[0056] Step S103. After bonding, the yellow and black cover films are cured. Multiple sensors on the production line collect process parameter data such as temperature, pressure, and position in real time. The collected process parameter data is analyzed in real time. When the process parameter data deviates from the preset range, an early warning signal is issued and the corresponding production equipment parameters are adjusted to ensure the manufacturing quality of the Mark point structure and the stability of the production process. Before making the Mark points, a machine learning-based intelligent algorithm is used to analyze and train the contrast between the cover film and the background copper surface and the SMT equipment recognition success rate of different color combinations in historical production data. A contrast-recognition success rate prediction model is established. Based on the contrast-recognition success rate prediction model, it is predicted whether the contrast between the yellow cover film and the background copper surface, and the white ink dots and the black cover film in the current design meet the recognition requirements of the SMT equipment. The cover film color parameters or Mark point design parameters are automatically adjusted according to the prediction results.

[0057] Specifically, the curing process involves heating and pressurizing the cover film adhesive to fully cure it, forming a stable laminated structure.

[0058] Real-time monitoring utilizes production line sensors to collect data such as temperature, pressure, and position, and combines this with machine learning algorithms to dynamically adjust process parameters, ensuring the quality and reliability of Mark points.

[0059] The curing process includes: Hot-press curing: The bonded PCB substrate is placed in a laminator or continuous curing oven, and a temperature gradient is set (e.g., heating rate 5-10℃ / min, peak temperature 180-220℃, holding time 60-90 minutes), with a pressure of 5-10MPa (adjusted according to the cover film thickness). Curing target: Complete cross-linking of the adhesive, and a bonding strength between the cover film and the substrate ≥5N / cm (peel test).

[0060] Sensor data acquisition includes: Temperature sensors: Thermocouples are placed inside the curing oven to monitor the temperature of each area in real time, with the error controlled within ±2℃. Pressure sensors: Pressure transmitters are installed on the laminator platen or bonding head to provide real-time feedback on bonding / curing pressure, with fluctuation range ≤ ±5% of the set value. Position sensors: The bonding position deviation of the cover film is monitored through a vision alignment system (such as a laser displacement sensor), with X / Y axis accuracy ≤ ±30μm.

[0061] The application of intelligent algorithms includes: Model training: Based on historical production data (such as the SMT recognition success rate corresponding to different yellow cover film color codes and white oil spot brightness values), a "contrast-recognition success rate" prediction model is established using machine learning algorithms (such as random forest and neural network). The input parameters include ΔE value, cover film transmittance, white oil spot reflectance, etc., and the output is the recognition success rate probability value.

[0062] During the Mark point design phase, input the contrast data of the current yellow overlay film and copper surface, and the white oil dots and black overlay film. The model predicts whether the recognition success rate is ≥99.5% (the threshold can be customized). If the prediction fails to meet the standard, the system automatically adjusts the parameters, including: Yellow overlay film: changing the color code (e.g., increasing yellow saturation) or adjusting the thickness (e.g., increasing from 25μm to 50μm to reduce light transmittance). White oil dots: adjusting the ink formula (e.g., increasing titanium dioxide content to improve brightness) or increasing the dot diameter (e.g., increasing from φ0.5mm to φ0.8mm). When the real-time collected process parameters (e.g., temperature exceeding limits, pressure fluctuations too large) deviate from the preset range, the system automatically issues an audible and visual warning and links the equipment to adjust parameters (e.g., adjusting heating power, pressure cylinder hydraulic pressure) to ensure process stability.

[0063] By eliminating drilling / laser cutting processes, equipment wear and processing time are saved (single-panel process efficiency is improved by more than 30%). The risk of Mark point deformation caused by window opening is avoided (the copper surface is protected throughout the pre-processing stage with a cover film). The immersion nickel-gold process is completely eliminated, reducing gold salt usage and material costs. White oil dots replace gold dots in the black cover film area, with ink costs only 1%-5% of gold salt costs.

[0064] By quantifying the relationship between color contrast and recognition success rate through machine learning, the traditional "trial and error" model is broken through, improving process reliability. The real-time closed-loop monitoring system reduces human intervention, improves production yield, and is suitable for large-scale automated production.

[0065] This method systematically solves the cost, efficiency, and reliability problems of traditional Mark point technology by combining material innovation, process simplification, and intelligent control. It is suitable for fields such as consumer electronics and automotive electronics that are sensitive to PCB precision and cost.

[0066] In some embodiments, the preparation of the PCB substrate and the pretreatment of the copper surface of the PCB substrate include: grinding the copper surface of the PCB substrate with a computer-controlled mechanical grinding device to remove the oxide layer and impurities on the copper surface, or degreasing and micro-etching the copper surface with a chemical cleaning line to make the copper surface reach a surface state suitable for subsequent lamination of the cover film.

[0067] The copper surface of the PCB substrate undergoes surface treatment, which involves mechanical grinding or chemical cleaning to remove the oxide layer and impurities, ensuring that the copper surface meets the bonding requirements.

[0068] Mechanical grinding uses computer-controlled equipment (such as a vertical grinding machine) equipped with nylon brushes or volcanic ash abrasive to mechanically grind the copper surface. The grinding speed is 5-10 m / min, the abrasive particle size is 80-120 mesh, and the grinding head pressure is adjusted in real time through a closed-loop control system to control the copper surface roughness at Ra 1.0-2.0 μm. At the same time, the data is fed back in real time by an online roughness detector.

[0069] The degreasing and micro-etching processes are performed via a chemical cleaning line. The degreasing section uses a 5%-10% sodium hydroxide solution at a temperature of 40-50℃ and a spray pressure of 0.2-0.3 MPa to remove grease. The micro-etching section uses a mixture of sulfuric acid (5%-8%) and hydrogen peroxide (3%-5%) at a temperature of 50-60℃, with an etching depth of 1-2 μm to remove the oxide layer. Throughout the process, a conductivity sensor monitors the solution concentration, and the solution is automatically replenished to maintain process stability.

[0070] In some embodiments, attaching the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area includes: using a computer-controlled vacuum pressing device to precisely attach the yellow cover film to the designated area of ​​the PCB substrate according to preset temperature, pressure and time parameters, and ensuring the accurate position of the cover film through the alignment system of the device during the pressing process.

[0071] The yellow covering film is precisely applied using a vacuum pressing device according to preset parameters to ensure accurate positioning.

[0072] Using computer-controlled vacuum laminating equipment (such as COG vacuum laminators), parameters such as laminating temperature (50-80℃), pressure (0.1-0.3MPa), and holding time (30-60 seconds) are pre-input into the equipment control system.

[0073] Before bonding, the CCD vision alignment system on the equipment identifies the pre-positioning marks (such as crosshairs or targets) on the PCB substrate and matches them with the alignment marks on the yellow cover film. The alignment accuracy is controlled within ±50μm.

[0074] The process employs a segmented lamination technique. First, a low pressure (0.05MPa) is used for pre-lamination to remove air bubbles. Then, the pressure is gradually increased to the set pressure to complete the lamination. Pressure fluctuations are monitored in real time by a pressure sensor to ensure that the cover film is flat and wrinkle-free.

[0075] In some embodiments, the cover film at the Mark point position is not subjected to windowing processing, including: during the design and manufacturing process of the cover film, the Mark point position is marked by computer-aided design software, processing data of the cover film is generated, the processing equipment is controlled not to perform drilling or laser cutting windowing operations at the Mark point position, and at the same time, a vision alignment system is used to ensure that the cover film at the Mark point position is accurately aligned with the target position on the PCB substrate.

[0076] By using a cover film Mark point location design to eliminate the need for window opening, the location is marked by CAD software and the processing equipment is controlled to avoid performing window opening operations, combined with visual alignment to ensure alignment.

[0077] In computer-aided design software (such as Altium Designer or CAM350), when designing the cover film layer, the Mark point positions (such as circular areas) are marked by layers, and the drilling or laser cutting instructions in that area are removed when generating Gerber data.

[0078] After the processing equipment (such as a laser cutting machine or CNC drilling machine) reads the design data, it automatically skips the Mark point position and does not perform window processing on that area.

[0079] During the bonding process, the relative position of the cover film and the substrate is monitored in real time by a visual alignment system (such as a laser displacement sensor). When the position of the Mark point is detected to be offset by more than ±30μm, the equipment automatically adjusts the position of the cover film to ensure that the cover film completely covers the copper surface area corresponding to the Mark point.

[0080] In some embodiments, the step of using the color of the yellow cover film itself to create a contrast with the background copper surface for SMT equipment recognition includes: before the cover film is applied, acquiring color images of the yellow cover film and the background copper surface through a computer image acquisition system, analyzing the contrast value of the two using an image processing algorithm, and confirming that the design meets the requirements when the contrast value reaches a preset threshold that the SMT equipment can recognize; otherwise, adjusting the cover film color or the copper surface processing technology.

[0081] The feasibility of identification was verified by quantifying the contrast between the yellow overlay film and the copper surface, and by image acquisition and algorithm analysis.

[0082] Before the cover film is applied, an industrial camera (such as a 12-megapixel CCD camera) with a ring light source is used to capture a stitched image of the yellow cover film and the pre-processed copper surface, with a resolution ≥1200 dpi. The color contrast between the two is analyzed using image processing algorithms (such as ΔE calculation based on the CIELAB color space). The calculation formula is: ΔE = (ΔL) / (ΔE) ) 2 +(Δa ) 2 +(Δb ) 2 Among them, the yellow covering film L a b Value and copper surface (approximately L) The difference between (a*=50, b*=20, b*=30) must satisfy ΔE≥15 (preset threshold).

[0083] If the contrast is not up to standard, the system will automatically prompt you to adjust the cover film color (e.g., replace it with YIELDER brand Y-1003 high saturation yellow film) or optimize the copper surface pretreatment micro-etching depth (e.g., increase it from 1.5μm to 2.0μm to deepen the copper surface color).

[0084] In some embodiments, for the black cover film area, firstly, white ink dots are made at the Mark point positions on the PCB substrate, including: using a computer-controlled inkjet printing device, according to the preset Mark point coordinates and size data, precisely spraying white ink at the Mark point positions on the PCB substrate to form white ink dots, and during the printing process, the position and shape of the ink dots are monitored in real time by the device's vision inspection system to ensure that they meet the design requirements.

[0085] White ink dots are created by using a black overlay film area, and ink is precisely ejected using a computer-controlled inkjet printer.

[0086] Using a piezoelectric inkjet printer, the Mark point coordinates (based on Gerber data conversion) and dimensional parameters (e.g., φ0.8mm circle) are imported into the device control software. The printhead movement speed is set to 10-20mm / s, the nozzle spacing to 50-100μm, and the jet pressure to 20-30kPa, ensuring that the white ink (e.g., Taiyo Ink 8200 series white ink) adheres evenly to the copper surface. During printing, the device's onboard vision inspection system (e.g., a line scan camera) scans the ink dots in real time, detecting positional deviations (≤±20μm) and shape accuracy (roundness error ≤5%). Defective dots automatically trigger a reprint mechanism. After printing, the ink is pre-baked at 80℃ for 15 minutes to fix its shape and prevent smudging during bonding.

[0087] In some embodiments, the step of attaching the black cover film to the corresponding position on the PCB substrate to expose the white ink dots includes: using a computer-controlled bonding device, identifying the position of the white ink dots on the PCB substrate using a vision alignment system, accurately aligning the pre-opened clearance window on the black cover film corresponding to the white ink dots with the white ink dots, and then bonding them to ensure that the white ink dots are fully exposed and the cover film is flat.

[0088] The black cover film is applied, and visual alignment is used to ensure that the white oil spots are fully exposed. The cover film avoids the window and precisely aligns with the white oil spots.

[0089] During the black coating film processing stage, a computer-controlled CNC punching machine or laser cutting machine is used to open avoidance windows in the non-mark point area according to the design data (the window size is 0.1-0.2mm larger than the white oil point), and no holes are punched at the mark point position.

[0090] High-precision bonding equipment is used during bonding. A dual-camera vision alignment system first identifies white ink dots on the PCB substrate and then aligns them with the clearance windows on the cover film, achieving an alignment accuracy of ≤±25μm. A vacuum adsorption platform is used to fix the substrate, and the cover film is gradually bonded at a pressure of 0.1MPa. A pressure sensor monitors the bonding stress distribution to ensure that there is no excess adhesive covering the edges of the white ink dots and that the surface flatness of the cover film is ≤5μm.

[0091] In some embodiments, after the bonding is completed, the yellow cover film and the black cover film are cured, including: placing the bonded PCB substrate into a computer-controlled curing oven, and performing heating, heat preservation and cooling treatments according to a preset temperature-time curve. During the curing process, the temperature data is monitored in real time by a temperature sensor in the oven, and the heating power is automatically adjusted by the control system to ensure that the cover film is completely cured and has stable performance.

[0092] The curing process for the cover film is carried out by executing a preset temperature-time curve in a computer-controlled curing oven, with temperature data monitored in real time.

[0093] The bonded PCB substrate is placed in a multi-stage curing oven (such as a Hotair Systems tunnel oven), with three temperature profiles set: Heating stage: heating from room temperature to 120℃ at a rate of 5℃ / min and maintaining for 10 minutes; Isothermal stage: heating to 180℃±5℃ and holding for 60 minutes to ensure complete cross-linking of the cover film adhesive; Cooling stage: natural cooling to room temperature at a rate ≤10℃ / min. 5-8 sets of thermocouple sensors are evenly arranged inside the oven to collect temperature data for each zone in real time. When the temperature in a certain area deviates from the set value by ±2℃, the control system automatically adjusts the heating power of the corresponding zone (e.g., using a PID control algorithm) to ensure temperature uniformity ≤±3℃. After curing, the adhesion is verified by a peel strength test, requiring a peel strength ≥5N / cm between the cover film and the copper surface. Defective products are automatically sorted to the rework line.

[0094] In some embodiments, the step of using machine learning-based intelligent algorithms to analyze and train the contrast of cover film and background copper surface with different color combinations and the SMT equipment recognition success rate in historical production data to establish a contrast-recognition success rate prediction model includes: acquiring cover film color parameters, background copper surface processing parameters, actual measured contrast values, and corresponding SMT equipment recognition success rate data stored in the historical production database through a computer data management system; cleaning and normalizing the acquired data using a computer preprocessing module to remove outlier data points and unify the data format; constructing a training dataset by using a computer feature engineering module with cover film color parameters, background copper surface processing parameters, and contrast values ​​as input features and SMT equipment recognition success rate as the output label; selecting a regression model or classification model from a preset machine learning algorithm library using a computer algorithm selection module, and optimizing the model parameters of the regression model or classification model through cross-validation; and iteratively training the selected regression model or classification model using a computer training module until the prediction error of the regression model or classification model on the test set is lower than a preset threshold, thereby obtaining the contrast-recognition success rate prediction model.

[0095] A contrast-based recognition success rate prediction model is built based on machine learning, including data acquisition, cleaning, feature engineering, model training and validation.

[0096] Data acquisition involves extracting historical production data from the MES (Manufacturing Execution System) through a computer data management system. This data includes cover film type, copper surface roughness (Ra value), white oil spot reflectance (measured by a spectrophotometer), measured contrast ΔE value, and the SMT recognition success rate of the corresponding batch (defined as the percentage of correct recognitions out of 1000 consecutive recognitions, based on AOI inspection statistics).

[0097] Data cleaning uses a preprocessing module to remove outlier data with ΔE < 10 or a recognition success rate < 90%. Numerical features (such as Ra and reflectivity) are normalized to the [-1, 1] interval using Z-score standardization. Feature engineering constructs an input feature vector: [Covering film color Lab value, Copper surface Lab value, Ra value, White oil spot Lab* value], with the output label being the recognition success rate (continuous values ​​use a regression model, and binary values ​​use a classification model).

[0098] Model training selects either a random forest regression model or a logistic regression classification model from the algorithm library, and uses 5-fold cross-validation to optimize parameters (such as the number of trees and maximum depth). Taking the regression model as an example, the training objective is to achieve a mean squared error (MSE) of <0.01 on the test set and a prediction accuracy ≥95%.

[0099] The model application automatically outputs a predicted recognition success rate by inputting the color parameters of the current cover film and copper surface in the new design. If the success rate is less than 99%, it triggers parameter adjustment suggestions (such as increasing the diameter of the white oil dots by 0.2mm or replacing the yellow film with a model of ΔE+3), and incorporates the new data into the training set to continuously optimize the model.

[0100] In some embodiments, to address the problem that traditional copper surface pretreatment (grinding / chemical cleaning) relies on fixed parameters and is difficult to adapt to fluctuations in copper surface conditions (such as oxide layer thickness and impurity distribution differences), reinforcement learning (RL) is introduced to construct an adaptive optimization framework. The pretreatment parameters are adjusted in real time by an agent to maximize the surface quality and contrast performance of subsequent overlay film bonding.

[0101] State perception is achieved by deploying high-resolution visual sensors (to acquire images of the copper oxide layer), roughness meters (to measure the Ra value of the copper surface), and electrochemical sensors (to detect the activity of the copper surface) in the pre-processing step of S101. This allows for the real-time acquisition of copper surface state characteristics (such as oxide layer thickness distribution, roughness, and impurity content), which serve as the state input for the RL agent.

[0102] The action space is defined by pre-processing parameters (such as the pressure (0.1-0.5MPa) and speed (1-5m / min) of mechanical grinding plates; the degreasing time (1-5min) and micro-etching solution concentration (10-30g / L) of chemical cleaning) as the action space of the agent, and each action corresponds to a set of parameter combinations.

[0103] The reward function is designed with "the surface smoothness after subsequent cover film lamination (evaluated by the visual inspection system)" and "the contrast between the yellow cover film and the copper surface (calculated by the image processing algorithm)" as the core reward indicators. Positive rewards (e.g., +5 points for a 10% increase in contrast) and negative penalties (e.g., -3 points for exceeding the surface smoothness standard) are set to guide the agent to learn the optimal parameter strategy.

[0104] A pre-trained RL model (such as the PPO algorithm) is generated using historical pre-processing data (copper surface condition → parameters → results), and then online learning is performed on the production line (the model is updated after each batch). During production, the agent outputs optimal parameters based on the real-time copper surface condition, controlling the grinding equipment or chemical cleaning line to automatically adjust. For example, when a thick oxide layer is detected on the copper surface, the grinding pressure is increased to 0.4 MPa and the micro-etching time is extended to 4 minutes to ensure that the oxide layer is completely removed and the copper surface roughness meets the bonding requirements.

[0105] Breaking through the limitations of traditional fixed parameters, reinforcement learning is used to achieve dynamic matching between preprocessing parameters and copper surface state, improving the consistency of cover film bonding and laying the foundation for subsequent Mark point contrast.

[0106] In some embodiments, to address the problem that traditional yellow cover film contrast relies on "static detection before bonding" and is difficult to handle color deviations during bonding (such as color difference between batches of cover film and color changes caused by bonding temperature), deep learning (DL) is introduced to build an end-to-end contrast prediction model, and combined with a closed-loop control system to adjust the cover film color or copper surface processing process in real time to ensure SMT recognition requirements.

[0107] Data acquisition and model training involved collecting historical production data, including color parameters of the yellow covering film (Lab). * The values ​​are measured by a spectrophotometer, copper surface treatment process parameters (such as micro-etching degree, PCB roughness), bonding process parameters (temperature, pressure), and contrast value after bonding (feedback from the SMT equipment identification system).

[0108] A convolutional neural network (CNN) model was constructed: It takes the overlay color image (captured by an industrial camera) and the copper surface state image (after preprocessing) as input, and outputs the predicted contrast value after lamination. The model was trained using transfer learning (fine-tuned based on ResNet-50 pre-trained on ImageNet) to improve generalization ability with small sample sizes.

[0109] Before the yellow cover film is laminated in S102, the contrast between the current cover film and the copper surface is predicted using a CNN model. If the predicted contrast does not reach the minimum recognizable threshold of the SMT equipment (e.g., ΔE≥30), a closed-loop adjustment is triggered: if the cover film color deviates (e.g., the yellow is too light), the color parameters of the cover film are automatically adjusted using computer-aided design (CAD) software (e.g., increasing the yellow ink concentration, adjusting the L* value from 80 to 75), and the result is sent to the cover film supplier to modify the production batch; if the copper surface treatment is insufficient (e.g., insufficient roughness), feedback is sent to the pre-processing stage in S101 to adjust the grinding parameters (e.g., increasing the speed to 4m / min) and reprocess the copper surface. After adjustment, the contrast is predicted again using the CNN model until the threshold requirement is met, and then the cover film lamination is performed.

[0110] The contrast prediction is moved from "post-lamination detection" to "pre-lamination prediction", and the dynamic optimization of the cover film color and copper surface treatment is achieved through deep learning and closed-loop control, avoiding rework due to substandard contrast.

[0111] In some embodiments, to address the problem that traditional black overlay white dots (fixed size / shape) are difficult to maximize the contrast with the black film, a Generative Adversarial Network (GAN) is introduced to generate the optimal white dot design scheme (size, shape, edge curvature). The effectiveness of the design is evaluated by a discriminator to ensure that the contrast between the white dots and the black film reaches the optimal state for SMT recognition.

[0112] The GAN model construction includes: Generator: Employing a U-Net structure, the input is parameters of the black overlay film (e.g., thickness (25-50μm), material (PI / PE), background copper color), and the output is the design parameters of the white dots (diameter (0.5-2mm), shape (circular / square / elliptical), and edge radius (0-0.5mm)). Discriminator: Employing a Convolutional Neural Network (CNN), the input is a simulated image of "white dots + black overlay film" (computer-generated), and the output is a contrast score (0-1, 1 indicating optimal), based on the actual recognition success rate of the SMT equipment (annotated using historical data).

[0113] The model training aims to maximize the contrast score given by the discriminator. The Wasserstein GAN with Gradient Penalty (WGAN-GP) algorithm is used to train both the generator and discriminator. Adversarial learning is employed to teach the generator to output white dot designs that conform to the characteristics of the black film. Training data consists of historical white dot designs (size, shape) and their corresponding SMT recognition success rates, as well as computer-simulated "white dot-black film" contrast data (calculated using image processing algorithms).

[0114] In the processing of the black overlay film area in S102, the parameters of the current black film (such as thickness 35μm, material PI) are first input into the generator to generate the optimal white oil dot design (such as diameter 1.2mm, elliptical shape, and corner radius 0.3mm).

[0115] Computer-aided manufacturing (CAM) software converts design parameters into inkjet printing instructions, controlling the inkjet equipment to precisely print white ink dots on the PCB substrate. After printing, the design validity is verified by a discriminator's simulated image scoring (e.g., score ≥ 0.9), ensuring optimal contrast between the white ink dots and the black film.

[0116] Breaking through the limitations of traditional "experience-based design" of white ink dots, this paper uses GAN to generate an optimal design that conforms to the characteristics of black film, thereby improving the success rate of SMT recognition and avoiding ink waste caused by over-design.

[0117] In some embodiments, to address the problem that traditional contrast-recognition success rate models (such as the regression model in the original S103) rely on single-factory data and are difficult to adapt to differences in equipment / materials across factories (such as the recognition threshold of SMT equipment in different factories and differences in cover film suppliers), federated learning (FL) is introduced to build a cross-factory collaborative training framework, which improves the generalization ability of the model without sharing the original data.

[0118] The federated learning framework is built by defining a central server (responsible for model aggregation) and edge nodes (local servers in each factory). The central server initializes a global contrast-to-recognition success rate model (such as a regression model based on XGBoost) and sends the model parameters to each edge node.

[0119] Each edge node (factory) uses local historical data (cover film color, copper surface processing parameters, contrast, SMT recognition success rate) to train a global model. During the training process, the original data is not uploaded, and only the updates of model parameters (such as gradient and weight) are calculated.

[0120] For example, the local data of factory A includes "yellow covering film (L*=85) + copper surface roughness (Ra=0.2μm) → contrast = 35 → recognition success rate = 98%", and the local data of factory B includes "yellow covering film (L*=80) + copper surface roughness (Ra=0.3μm) → contrast = 32 → recognition success rate = 95%". Both factories train their local models separately.

[0121] Each edge node sends its local model parameter updates to the central server. The server uses a weighted average strategy (the weights are determined by the amount of data from each factory; for example, if factory A accounts for 60% of the data, its weight is 0.6) to aggregate the parameters and generate a new global model.

[0122] The central server sends the global model back to each edge node and repeats the "local training-aggregation" process (e.g., 10 rounds) until the prediction error of the model on the validation set of each factory is lower than the preset threshold (e.g., MAE≤2%).

[0123] In the Mark point design verification of S103, each factory used a global model optimized by federated learning to predict the contrast and recognition success rate of the current design (such as yellow cover film L*=82, copper surface Ra=0.25μm).

[0124] If the predicted recognition success rate does not meet the factory's SMT equipment requirements (e.g., ≥99%), the model will combine the factory's local data (e.g., the historical recognition success rate of "L*=83, Ra=0.25μm" was 99.5%) and recommend adjusting the cover film color (e.g., increasing L* from 82 to 83) or the copper surface treatment process (e.g., reducing Ra to 0.2μm).

[0125] By enabling collaborative optimization of cross-factory models through federated learning, the data privacy of each factory is protected, and the model's generalization ability to different equipment / materials is improved, thus solving the problem of traditional models "failing when changing factories".

[0126] In some embodiments, to address the problems of traditional process monitoring relying on "threshold judgment" (such as temperature deviation ±5℃ alarm), difficulty in identifying minor anomalies (such as temperature gradient fluctuations, slow pressure drift), and inability to automatically adjust parameters, a framework combining anomaly detection algorithms (such as Autoencoder) and reinforcement learning is introduced to achieve real-time anomaly identification and intelligent parameter adjustment, thereby improving production stability.

[0127] The anomaly detection module uses a variational autoencoder (VAE) to build an anomaly detection model. The input is the multi-dimensional parameters (temperature, pressure, position, curing time) monitored by the S103 process, and the output is the reconstruction error.

[0128] The training data comes from the process parameters of normal production batches (such as the curing oven temperature curve conforming to the preset and the pressure being stable), and the model learns the distribution of normal data. When the parameters collected in real time are input into the model, if the reconstruction error exceeds the abnormal threshold (such as the 95th percentile), it is judged as abnormal (such as the temperature gradient increasing from 1℃ / min to 3℃ / min).

[0129] The reinforcement learning adjustment module takes anomaly types (such as abnormal temperature gradients and pressure drift) as state inputs, adjustment actions (such as adjusting the heating power of the curing oven and the pressure of the pressing equipment) as action space, and "anomaly recovery time" and "product qualification rate" as reward indicators to construct an RL agent.

[0130] For example, when VAE detects "abnormal temperature gradient in curing oven (3℃ / min)," the agent's state is "abnormal temperature gradient," the action space is "increase heating power by 10%" and "decrease conveyor belt speed by 20%," and the reward function is "reward +10 points for abnormal recovery time ≤ 5min" and "reward +5 points for product qualification rate ≥ 99%."

[0131] In the curing / pressing process of S103, the VAE monitors process parameters in real time. If an anomaly is detected (such as an abnormal temperature gradient), the RL agent is triggered. The agent outputs the optimal adjustment action based on the anomaly type (such as reducing the conveyor belt speed by 20% to slow down the heating rate) and controls the production equipment to adjust parameters. After adjustment, the VAE continues to monitor parameters. If the reconstruction error returns to the normal range (anomaly recovery), the agent receives a positive reward; if not, the agent updates its strategy through online learning (such as attempting to increase the heating power by 15%).

[0132] The anomaly detection is upgraded from "passive alarm" to "active adjustment". The VAE identifies minor anomalies and then the RL agent automatically optimizes the parameters. This is more timely and accurate than the traditional threshold judgment, and reduces the scrap rate caused by anomalies (such as incomplete curing of the cover film).

[0133] This application provides a windowless Mark dot structure for PCB cover film, which is manufactured based on the method provided in any embodiment of this application.

[0134] In some embodiments, the provided windowless Mark point structure is applied to a PCB substrate, and its core structure consists of four parts: the copper surface of the PCB substrate, the Mark points in the yellow cover film area, the Mark points in the black cover film area, and the process monitoring module.

[0135] The copper surface of the PCB substrate serves as the background layer for the Mark points. The surface undergoes pretreatment (such as grinding and cleaning) to ensure roughness and adhesion, providing conditions for cover film lamination and ink adhesion.

[0136] The yellow overlay area contrasts with the metallic hue of the copper surface, creating a color difference; the white ink dots in the black overlay area are printed directly onto the copper surface, achieving marking through the color difference between the white ink and the black overlay.

[0137] The yellow cover film markings are applied directly and completely to the copper surface of the PCB substrate without drilling or laser cutting at the marked locations. Utilizing the transparent or semi-transparent properties of the yellow cover film, its color creates a natural color contrast with the gold / light red of the underlying copper surface (RGB contrast difference ≥80), facilitating detection by the visual recognition system of SMT equipment. This eliminates the need for windowing, reducing wear and tear on drilling / laser cutting equipment and processing time, thus lowering costs.

[0138] The cover film completely covers the Mark point area, preventing the copper surface Mark points from deforming due to exposure during pre-processing of the board, thus improving the reliability of identification.

[0139] The black overlay area Mark points are created by first printing circular or square white ink dots (white ink thickness 20-30μm, diameter 0.8-1.2mm) at the Mark point positions on the PCB substrate. The ink used is a special white photosensitive ink with high contrast and chemical corrosion resistance.

[0140] A black cover film is bonded to the substrate. No window is needed at the Mark point location on the cover film. The white dots are exposed through the transparency or ultra-thin nature of the cover film (thickness ≤ 50μm), creating a high-contrast combination between the white dots and the black cover film (RGB contrast difference between white and black ≥ 150). This replaces the traditional windowed immersion gold process, eliminating the need for the nickel-gold immersion step, reducing gold salt usage, and lowering material costs. The smooth surface of the white dots avoids SMT recognition errors caused by uneven reflection on the gold surface, improving recognition accuracy. Complete coverage by the cover film eliminates the need for custom-made punches in the die-cutting process, reducing mold costs by 20%.

[0141] The process monitoring module collects process parameters in real time (such as bonding temperature 150-180℃, pressure 0.5-1.0MPa, and alignment deviation ≤0.1mm) by deploying temperature sensors (accuracy ±0.5℃), pressure sensors (accuracy ±1%FS), and position vision sensors (accuracy ±0.05mm) during the cover film bonding and curing process.

[0142] A contrast-to-recognition success rate prediction model is trained based on historical production data (using a convolutional neural network CNN algorithm). The input parameters are the color parameters of the overlay film (e.g., yellow Lab value: L=80, a=-10, b=60; black Lab value: L=10, a=0, b=0) and the white oil spot parameters (e.g., reflectivity ≥85%). The output is the predicted recognition success rate (error ≤3%).

[0143] When the system detects that the process parameters deviate from the preset range or the predicted recognition success rate is less than 95%, it automatically adjusts the cover film bonding pressure, curing time, or ink printing thickness to ensure the consistency of the Mark point structure.

[0144] Please see Figure 2 As shown, Figure 2 This is a schematic diagram of a system 200 for fabricating a windowless Mark dot structure for a PCB cover film, provided in an embodiment of this application. This system 200 is used to perform the steps of the method for fabricating a windowless Mark dot structure for a PCB cover film 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, a laptop computer, a wearable device, or a robot.

[0145] like Figure 2 As shown, the fabrication system 200 for the windowless Mark dot structure of PCB cover film includes:

[0146] The copper surface processing unit 201 is used to prepare the PCB substrate and perform pretreatment on the copper surface of the PCB substrate.

[0147] The cover bonding unit 202 is used to bond the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area. The cover film at the Mark point position is not opened, and the color of the yellow cover film itself is used to form a contrast with the background copper surface for SMT equipment to identify. For the black cover film area, white ink dots are first made at the Mark point position on the PCB substrate, and the black cover film is bonded to the corresponding position on the PCB substrate to expose the white ink dots.

[0148] The bonding completion unit 203 is used to cure the yellow and black cover films after bonding. Multiple sensors deployed on the production line collect process parameter data such as temperature, pressure, and position in real time. The collected process parameter data is analyzed in real time. When the process parameter data deviates from the preset range, an early warning signal is issued and the corresponding production equipment parameters are adjusted to ensure the manufacturing quality of the Mark point structure and the stability of the production process. Before the Mark points are made, a machine learning-based intelligent algorithm is used to analyze and train the contrast between the cover film and the background copper surface and the SMT equipment recognition success rate of different color combinations in historical production data. A contrast-recognition success rate prediction model is established. Based on the contrast-recognition success rate prediction model, it is predicted whether the contrast between the yellow cover film and the background copper surface, and the white ink dots and the black cover film in the current design meet the recognition requirements of the SMT equipment. The cover film color parameters or Mark point design parameters are automatically adjusted according to the prediction results.

[0149] In some embodiments, the preparation of the PCB substrate and the pretreatment of the copper surface of the PCB substrate include: grinding the copper surface of the PCB substrate with a computer-controlled mechanical grinding device to remove the oxide layer and impurities on the copper surface, or degreasing and micro-etching the copper surface with a chemical cleaning line to make the copper surface reach a surface state suitable for subsequent lamination of the cover film.

[0150] In some embodiments, attaching the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area includes: using a computer-controlled vacuum pressing device to precisely attach the yellow cover film to the designated area of ​​the PCB substrate according to preset temperature, pressure and time parameters, and ensuring the accurate position of the cover film through the alignment system of the device during the pressing process.

[0151] In some embodiments, the cover film at the Mark point position is not subjected to windowing processing, including: during the design and manufacturing process of the cover film, the Mark point position is marked by computer-aided design software, processing data of the cover film is generated, the processing equipment is controlled not to perform drilling or laser cutting windowing operations at the Mark point position, and at the same time, a vision alignment system is used to ensure that the cover film at the Mark point position is accurately aligned with the target position on the PCB substrate.

[0152] In some embodiments, the step of using the color of the yellow cover film itself to create a contrast with the background copper surface for SMT equipment recognition includes: before the cover film is applied, acquiring color images of the yellow cover film and the background copper surface through a computer image acquisition system, analyzing the contrast value of the two using an image processing algorithm, and confirming that the design meets the requirements when the contrast value reaches a preset threshold that the SMT equipment can recognize; otherwise, adjusting the cover film color or the copper surface processing technology.

[0153] In some embodiments, for the black cover film area, firstly, white ink dots are made at the Mark point positions on the PCB substrate, including: using a computer-controlled inkjet printing device, according to the preset Mark point coordinates and size data, precisely spraying white ink at the Mark point positions on the PCB substrate to form white ink dots, and during the printing process, the position and shape of the ink dots are monitored in real time by the device's vision inspection system to ensure that they meet the design requirements.

[0154] In some embodiments, the step of attaching the black cover film to the corresponding position on the PCB substrate to expose the white ink dots includes: using a computer-controlled bonding device, identifying the position of the white ink dots on the PCB substrate using a vision alignment system, accurately aligning the pre-opened clearance window on the black cover film corresponding to the white ink dots with the white ink dots, and then bonding them to ensure that the white ink dots are fully exposed and the cover film is flat.

[0155] In some embodiments, after the bonding is completed, the yellow cover film and the black cover film are cured, including: placing the bonded PCB substrate into a computer-controlled curing oven, and performing heating, heat preservation and cooling treatments according to a preset temperature-time curve. During the curing process, the temperature data is monitored in real time by a temperature sensor in the oven, and the heating power is automatically adjusted by the control system to ensure that the cover film is completely cured and has stable performance.

[0156] In some embodiments, the step of using machine learning-based intelligent algorithms to analyze and train the contrast of cover film and background copper surface with different color combinations and the SMT equipment recognition success rate in historical production data to establish a contrast-recognition success rate prediction model includes: acquiring cover film color parameters, background copper surface processing parameters, actual measured contrast values, and corresponding SMT equipment recognition success rate data stored in the historical production database through a computer data management system; cleaning and normalizing the acquired data using a computer preprocessing module to remove outlier data points and unify the data format; constructing a training dataset by using a computer feature engineering module with cover film color parameters, background copper surface processing parameters, and contrast values ​​as input features and SMT equipment recognition success rate as the output label; selecting a regression model or classification model from a preset machine learning algorithm library using a computer algorithm selection module, and optimizing the model parameters of the regression model or classification model through cross-validation; and iteratively training the selected regression model or classification model using a computer training module until the prediction error of the regression model or classification model on the test set is lower than a preset threshold, thereby obtaining the contrast-recognition success rate prediction model.

[0157] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the above-described system for fabricating the windowless Mark dot structure for PCB cover film and each module can be referred to the corresponding content in the various embodiments of the above-described method for fabricating the windowless Mark dot structure for PCB cover film, and will not be repeated here.

[0158] The above-described method for fabricating the windowless Mark dot structure for PCB cover film can be implemented as a computer program, which can be used in various ways, such as... Figure 2 It runs on the device shown.

[0159] 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.

[0160] 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 for fabricating a windowless Mark dot structure for a PCB cover film.

[0161] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0162] 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 execute any method for manufacturing windowless Mark dot structures for PCB cover films.

[0163] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 3 The 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.

[0164] 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.

[0165] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps:

[0166] Prepare the PCB substrate and perform pretreatment on the copper surface of the PCB substrate.

[0167] For the yellow cover film area of ​​the PCB substrate, the yellow cover film is attached to the corresponding position of the PCB substrate. The cover film at the Mark point position is not opened. The color of the yellow cover film itself is used to form a contrast with the background copper surface for SMT equipment to identify. For the black cover film area, white ink dots are first made at the Mark point position of the PCB substrate. The black cover film is then attached to the corresponding position of the PCB substrate to expose the white ink dots.

[0168] After lamination, the yellow and black cover films are cured. Multiple sensors deployed on the production line collect real-time process parameter data on temperature, pressure, and position. This data is analyzed in real time, and when deviations from preset ranges are detected, an early warning signal is issued, and corresponding production equipment parameters are adjusted to ensure the quality of the Mark point structure and the stability of the production process. Before creating the Mark points, a machine learning-based intelligent algorithm is used to analyze and train the contrast between the cover film and the background copper surface for different color combinations and the SMT equipment recognition success rate in historical production data. A contrast-recognition success rate prediction model is established. Based on this model, it is predicted whether the contrast between the yellow cover film and the background copper surface, and between the white ink dots and the black cover film in the current design meets the recognition requirements of the SMT equipment. The cover film color parameters or Mark point design parameters are automatically adjusted based on the prediction results.

[0169] In some embodiments, the preparation of the PCB substrate and the pretreatment of the copper surface of the PCB substrate include: grinding the copper surface of the PCB substrate with a computer-controlled mechanical grinding device to remove the oxide layer and impurities on the copper surface, or degreasing and micro-etching the copper surface with a chemical cleaning line to make the copper surface reach a surface state suitable for subsequent lamination of the cover film.

[0170] In some embodiments, attaching the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area includes: using a computer-controlled vacuum pressing device to precisely attach the yellow cover film to the designated area of ​​the PCB substrate according to preset temperature, pressure and time parameters, and ensuring the accurate position of the cover film through the alignment system of the device during the pressing process.

[0171] In some embodiments, the cover film at the Mark point position is not subjected to windowing processing, including: during the design and manufacturing process of the cover film, the Mark point position is marked by computer-aided design software, processing data of the cover film is generated, the processing equipment is controlled not to perform drilling or laser cutting windowing operations at the Mark point position, and at the same time, a vision alignment system is used to ensure that the cover film at the Mark point position is accurately aligned with the target position on the PCB substrate.

[0172] In some embodiments, the step of using the color of the yellow cover film itself to create a contrast with the background copper surface for SMT equipment recognition includes: before the cover film is applied, acquiring color images of the yellow cover film and the background copper surface through a computer image acquisition system, analyzing the contrast value of the two using an image processing algorithm, and confirming that the design meets the requirements when the contrast value reaches a preset threshold that the SMT equipment can recognize; otherwise, adjusting the cover film color or the copper surface processing technology.

[0173] In some embodiments, for the black cover film area, firstly, white ink dots are made at the Mark point positions on the PCB substrate, including: using a computer-controlled inkjet printing device, according to the preset Mark point coordinates and size data, precisely spraying white ink at the Mark point positions on the PCB substrate to form white ink dots, and during the printing process, the position and shape of the ink dots are monitored in real time by the device's vision inspection system to ensure that they meet the design requirements.

[0174] In some embodiments, the step of attaching the black cover film to the corresponding position on the PCB substrate to expose the white ink dots includes: using a computer-controlled bonding device, identifying the position of the white ink dots on the PCB substrate using a vision alignment system, accurately aligning the pre-opened clearance window on the black cover film corresponding to the white ink dots with the white ink dots, and then bonding them to ensure that the white ink dots are fully exposed and the cover film is flat.

[0175] In some embodiments, after the bonding is completed, the yellow cover film and the black cover film are cured, including: placing the bonded PCB substrate into a computer-controlled curing oven, and performing heating, heat preservation and cooling treatments according to a preset temperature-time curve. During the curing process, the temperature data is monitored in real time by a temperature sensor in the oven, and the heating power is automatically adjusted by the control system to ensure that the cover film is completely cured and has stable performance.

[0176] In some embodiments, the step of using machine learning-based intelligent algorithms to analyze and train the contrast of cover film and background copper surface with different color combinations and the SMT equipment recognition success rate in historical production data to establish a contrast-recognition success rate prediction model includes: acquiring cover film color parameters, background copper surface processing parameters, actual measured contrast values, and corresponding SMT equipment recognition success rate data stored in the historical production database through a computer data management system; cleaning and normalizing the acquired data using a computer preprocessing module to remove outlier data points and unify the data format; constructing a training dataset by using a computer feature engineering module with cover film color parameters, background copper surface processing parameters, and contrast values ​​as input features and SMT equipment recognition success rate as the output label; selecting a regression model or classification model from a preset machine learning algorithm library using a computer algorithm selection module, and optimizing the model parameters of the regression model or classification model through cross-validation; and iteratively training the selected regression model or classification model using a computer training module until the prediction error of the regression model or classification model on the test set is lower than a preset threshold, thereby obtaining the contrast-recognition success rate prediction model.

[0177] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the steps of the method for manufacturing a windowless Mark dot structure for a PCB cover film as provided in any embodiment of this application.

[0178] 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.

[0179] 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 fabricating a windowless Mark dot structure for PCB cover film, characterized in that, The method includes: Prepare the PCB substrate and perform pretreatment on the copper surface of the PCB substrate. For the yellow cover film area of ​​the PCB substrate, the yellow cover film is attached to the corresponding position of the PCB substrate. The cover film at the Mark point position is not opened. The color of the yellow cover film itself is used to form a contrast with the background copper surface for SMT equipment to identify. For the black cover film area, white ink dots are first made at the Mark point position of the PCB substrate. The black cover film is then attached to the corresponding position of the PCB substrate to expose the white ink dots. After lamination, the yellow and black cover films are cured. Multiple sensors deployed on the production line collect real-time process parameter data on temperature, pressure, and position. This data is analyzed in real time, and when deviations from preset ranges are detected, an early warning signal is issued, and corresponding production equipment parameters are adjusted to ensure the quality of the Mark point structure and the stability of the production process. Before creating the Mark points, a machine learning-based intelligent algorithm is used to analyze and train the contrast between the cover film and the background copper surface for different color combinations and the SMT equipment recognition success rate in historical production data. A contrast-recognition success rate prediction model is established. Based on this model, it is predicted whether the contrast between the yellow cover film and the background copper surface, and between the white ink dots and the black cover film in the current design meets the recognition requirements of the SMT equipment. The cover film color parameters or Mark point design parameters are automatically adjusted based on the prediction results.

2. The method according to claim 1, characterized in that, The preparation of the PCB substrate and the pretreatment of the copper surface of the PCB substrate include: The copper surface of the PCB substrate is ground by computer-controlled mechanical grinding equipment to remove the oxide layer and impurities, or the copper surface is degreased and micro-etched by chemical cleaning line to make the copper surface suitable for subsequent bonding of cover film.

3. The method according to claim 1, characterized in that, The process of attaching the yellow cover film to the corresponding position on the PCB substrate for the yellow cover film area includes: Using computer-controlled vacuum lamination equipment, the yellow cover film is precisely adhered to the designated area of ​​the PCB substrate according to preset temperature, pressure and time parameters. During the lamination process, the alignment system of the equipment ensures that the cover film is in an accurate position.

4. The method according to claim 1, characterized in that, The cover film at the Mark point is not windowed, including: During the design and manufacturing of the cover film, computer-aided design software is used to mark the Mark point positions, generate processing data for the cover film, and control the processing equipment to prevent drilling or laser cutting at the Mark point positions. At the same time, a vision alignment system is used to ensure that the cover film is precisely aligned with the target position on the PCB substrate at the Mark point positions.

5. The method according to claim 1, characterized in that, The use of yellow overlay film to create contrast between its color and the background copper surface for SMT equipment recognition includes: Before the cover film is applied, the color images of the yellow cover film and the background copper surface are acquired by a computer image acquisition system. The contrast value of the two is analyzed by the image processing algorithm. When the contrast value reaches the preset threshold that the SMT equipment can recognize, it is confirmed whether the contrast between the yellow cover film and the background copper surface, and between the white ink dots and the black cover film in the current design meets the recognition requirements of the SMT equipment. Otherwise, the cover film color or the copper surface processing technology is adjusted.

6. The method according to claim 1, characterized in that, For the black overlay area, white ink dots are first created at the Mark points on the PCB substrate, including: Using a computer-controlled inkjet printer, white ink is precisely sprayed onto the Mark point positions on the PCB substrate according to the pre-set Mark point coordinates and size data, forming white ink dots. During the printing process, the position and shape of the ink dots are monitored in real time by the equipment's vision inspection system.

7. The method according to claim 1, characterized in that, The step of attaching the black cover film to the corresponding position on the PCB substrate to expose the white ink dots includes: Using a computer-controlled bonding device, a vision alignment system is used to identify the position of white ink dots on the PCB substrate. The pre-cut clearance windows on the black cover film corresponding to the white ink dots are precisely aligned with the white ink dots before bonding, ensuring that the white ink dots are fully exposed and the cover film is bonded smoothly.

8. The method according to claim 1, characterized in that, After the bonding is completed, the yellow and black cover films undergo a curing process, including: The bonded PCB substrate is placed in a computer-controlled curing oven, where it is heated, kept warm, and cooled according to a preset temperature-time curve. During the curing process, the temperature data is monitored in real time by a temperature sensor inside the oven, and the heating power is automatically adjusted by the control system to ensure that the cover film is completely cured and has stable performance.

9. The method according to claim 1, characterized in that, The method utilizes machine learning-based intelligent algorithms to analyze and train the contrast between the cover film and the background copper surface with different color combinations in historical production data, as well as the recognition success rate of SMT equipment, to establish a contrast-recognition success rate prediction model, including: The computer data management system retrieves the cover film color parameters, background copper surface processing parameters, actual measured contrast values, and corresponding SMT equipment recognition success rate data stored in the historical production database. The computer's preprocessing module is used to clean and normalize the acquired data, remove outlier data points, and unify the data format. Using the feature engineering module of the computer, the cover film color parameters, background copper surface processing parameters and contrast values ​​are used as input features, and the SMT equipment recognition success rate is used as the output label to construct a training dataset. The computer's algorithm selection module selects a regression model or classification model from a pre-set machine learning algorithm library, and optimizes the model parameters of the regression model or classification model through cross-validation. The computer's training module iteratively trains the selected regression model or classification model until the prediction error of the regression model or classification model on the test set is lower than a preset threshold, thereby obtaining a contrast-recognition success rate prediction model.

10. A windowless Mark dot structure for PCB cover film, characterized in that, It is made according to the method of any one of claims 1-9.

Images