Manufacturing process of anti-plating PCB (Printed Circuit Board) gold-plated board
Through the U-Net model, the mask layer defects are identified and combined with the use of spin coating machines and photosensitive masking agents, the penetration and unevenness of the gold-plated layer caused by mask layer defects during the gold plating process are solved, and the high quality of the gold-plated layer is achieved and Uniformity.
Patent Information
- Application Number
- CN202411188560.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-28
- Publication Date
- 2024-11-29
AI Technical Summary
In the prior art, anti-seepage treatment is prone to mask layer defects, such as bubbles and wrinkles, during the gold plating process, resulting in penetration or unevenness of the gold plating layer, and these defects are difficult to identify.
The U-Net model is used to analyze the mask layer image, and the defects in the mask layer are identified through the convolution layer, activation function, pooling layer, bridge part and decoding part, and combined with the spin coating machine and a specific photosensitive masking agent for prevention. The seepage treatment ensures the uniformity of the gold-plated layer.
Effectively identify and avoid defective mask layers entering the gold plating process, ensure the quality and uniformity of the gold plating layer, and reduce the penetration and unevenness of the gold plating layer.
Smart Images

Figure CN119053047A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of PCB gold plating, and in particular relates to a manufacturing process of a PCB gold-plated plate with anti-seepage plating. Background Art
[0002] In PCB (Printed Circuit Board) manufacturing, gold plating is a common surface treatment method, which is usually used to improve the reliability and corrosion resistance of electrical contacts. During the gold plating process, anti-seepage treatment is required to prevent the metal plating from penetrating into other areas or layers of the board and ensure the quality and uniformity of the plating.
[0003] At present, anti-seepage treatment can be achieved by coating a mask layer, that is, before gold plating, a mask layer is usually coated on the surface of the PCB to prevent the gold plating layer from penetrating into the non-contact area. This mask layer can be a photoresist or other types of protective coatings; however, defects such as bubbles and wrinkles may occur in the mask layer during the coating process, and some tiny bubbles or wrinkles cannot be manually identified, and flowing into the subsequent gold plating process will cause the gold plating layer to penetrate or become uneven; therefore, a manufacturing process for PCB gold-plated boards with anti-seepage plating is proposed. Summary of the invention
[0004] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a manufacturing process for a PCB gold-plated plate with anti-seepage plating, which solves the problems in the prior art.
[0005] The purpose of the present invention can be achieved through the following technical solutions:
[0006] A process for manufacturing a gold-plated PCB with an anti-seepage plating comprises the following steps:
[0007] S1, cleaning the PCB and coating a photosensitive masking agent on the surface of the PCB to form a masking layer;
[0008] S2, collects the mask layer image and uses the U-Net model to analyze the mask layer image to achieve defect recognition and detection;
[0009] S3, performing gold plating on the PCB that has been detected to have no mask layer defects to obtain a gold-plated PCB;
[0010] By introducing the U-Net model in the process of mask layer defect detection to analyze the image and identify defects such as bubbles and wrinkles in the mask layer, PCBs with defective mask layers can be prevented from flowing into the gold plating process, causing penetration or unevenness of the gold plating layer.
[0011] Furthermore, the photosensitive masking agent is: polyimide, photosensitive resin, phenolic resin, photosensitive polyurethane or photosensitive silicone rubber.
[0012] Furthermore, a spin coater is used for coating the photosensitive mask agent, the spin coating speed is 500-600 r / min, and the thickness of the mask layer is 3-8 um.
[0013] Furthermore, the U-Net model includes: an encoding part, a bridging part and a decoding part;
[0014] The encoding part includes:
[0015] Convolutional layer: extracts image features through a series of convolution operations;
[0016] Activation function: Use ReLU activation function;
[0017] Pooling layer: Use maximum pooling to reduce the spatial dimension of the image;
[0018] The bridge part includes a convolution layer to process high-dimensional features and extract complex features;
[0019] The decoding part includes:
[0020] Upsampling layer: restores the spatial resolution of the image through transposed convolution or upsampling operations;
[0021] Convolutional layer: Connects to the image features of the encoded part, fuses low-level features and high-level features through convolution operations, and generates a predicted segmentation map of the mask layer.
[0022] Furthermore, the steps of using the U-Net model to perform defect detection on the mask layer image include:
[0023] S21, collecting an image of the mask layer using a CCD camera or a microscope camera;
[0024] S22, preprocessing the collected mask layer image;
[0025] S23, input the preprocessed image into the U-Net model for training;
[0026] S24, using the trained U-Net model to segment the mask layer image and output defect recognition results.
[0027] Furthermore, the preprocessing of the mask layer image includes: normalization processing and data enhancement processing;
[0028] The normalization formula is:
[0029]
[0030] Where I is the original pixel value; I min and I max are the minimum and maximum pixel values in the image respectively; I normis the normalized pixel value in the range of [-1, 1].
[0031] Furthermore, the data enhancement processing includes: rotation, translation and scaling of the image;
[0032] The formula for image rotation is:
[0033]
[0034] Where (x, y) is the original coordinate, (x′, y′) is the rotated coordinate; θ is the rotation angle;
[0035] The image translation formula is:
[0036]
[0037] Where (x″, y″) is the coordinate after translation; Δx and Δy are the translation amounts in the x and y directions;
[0038] The image scaling formula is:
[0039]
[0040] Wherein, (x″′, y″′) is the scaled coordinate; s is the scaling factor.
[0041] Furthermore, the loss function of the U-Net model is:
[0042]
[0043] Among them, y i is the true label, p i is the predicted probability.
[0044] Furthermore, the steps of gold plating the PCB include:
[0045] S31, performing acid etching on the mask layer;
[0046] S32, gold-plating the PCB using electroplating gold process;
[0047] S33, cleaning the gold-plated layer, and drying it to obtain a gold-plated PCB.
[0048] A system for manufacturing a PCB gold-plated plate with anti-seepage plating, comprising:
[0049] Mask layer coating module: cleans the PCB and applies a photosensitive masking agent on the PCB surface to form a mask layer;
[0050] Image analysis and detection module: collects mask layer images and uses the U-Net model to analyze the mask layer images to achieve defect recognition and detection;
[0051] And, a gold-plating module: gold-plating is performed on the PCB that has been detected to have no mask layer defects to obtain a gold-plated PCB board.
[0052] Beneficial effects of the present invention:
[0053] 1. Prevent metal penetration by selecting a suitable photosensitive masking agent for coating before gold plating; and use low-speed spin coating to reduce bubbles and ensure uniform coverage.
[0054] 2. By introducing the U-Net model in the process of mask layer defect detection to analyze the image and identify defects such as bubbles and wrinkles in the mask layer, PCBs with defective mask layers are prevented from flowing into the gold plating process, causing penetration or unevenness of the gold plating layer. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0056] Figure 1 The present invention is a schematic diagram of the manufacturing process of the PCB gold-plated board. DETAILED DESCRIPTION
[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0058] like Figure 1 As shown, a process for manufacturing a gold-plated PCB with anti-seepage plating includes the following steps:
[0059] S1, coating a mask layer on the PCB surface;
[0060] Specifically include:
[0061] S11, clean the PCB surface to ensure that there is no oil or dust on the surface;
[0062] S12, dropping a photosensitive masking agent at the center of the PCB, and coating the photosensitive masking agent on the surface of the PCB using a spin coater to form a mask layer;
[0063] The photosensitive masking agent can be specifically selected from: polyimide, photosensitive resin, phenolic resin, photosensitive polyurethane or photosensitive silicone rubber; among them, polyimide masking agent can remain stable under high temperature and harsh chemical environment, and is suitable for complex PCB manufacturing processes, including gold plating. Photosensitive resin can form the required mask pattern according to the development reaction after exposure. Phenolic resin is usually used to make photosensitive mask layers, providing excellent resolution and edge clarity, suitable for precise mask layer requirements. Photosensitive polyurethane masking agent has high resolution and good lithography performance, can provide clear mask patterns in precision manufacturing processes, and is resistant to chemical corrosion. Photosensitive silicone rubber has excellent high temperature resistance and chemical stability, suitable for complex PCB manufacturing processes and high temperature processing environments.
[0064] Among them, the spin coating speed of the spin coater is 500-600r / min, which is a low-speed spin coating, suitable for reducing bubbles and ensuring uniform coating of the mask layer. Slow-speed spin coating allows the masking agent to have more time to be evenly distributed on the PCB surface and reduce the generation of bubbles.
[0065] In addition, the thickness of the mask layer after coating is 3-8um. The required thickness can be achieved by selecting the appropriate spin coating speed and time according to the viscosity of the mask agent and the spin coating parameters. The ambient temperature is 20-25℃, which is conducive to the uniform coating and drying of the mask agent. The relative humidity is controlled between 40-60% to reduce bubbles or uneven drying of the mask layer.
[0066] S2, performing defect detection on the mask layer after coating;
[0067] The specific steps include:
[0068] S21, using a high-resolution camera (such as a CCD camera or a microscope camera) to collect an image of the mask layer;
[0069] S22, preprocessing the collected mask layer image;
[0070] Preprocessing specifically includes: normalization processing and data enhancement processing;
[0071] Normalization is to scale the pixel values to a standard range so that the pixel values of different images are at the same scale, which enhances the adaptability of the model to the input data, avoids the gradient explosion or gradient vanishing problem caused by the input data range being too large or too small, helps to accelerate the training process of the model, and improves the stability and accuracy of the model. In this embodiment, the formula for normalization is:
[0072]
[0073] Where I is the original pixel value; I min and I maxare the minimum and maximum pixel values in the image respectively; I norm is the normalized pixel value in the range of [-1, 1].
[0074] Data augmentation processing is to increase the diversity of training data by performing various transformations on images, thereby improving the generalization ability of the model; data augmentation processing specifically includes: image rotation, translation and scaling;
[0075] The formula for image rotation is:
[0076]
[0077] Where (x, y) is the original coordinate, (x′, y′) is the rotated coordinate; θ is the rotation angle (expressed in radians);
[0078] The image translation formula is:
[0079]
[0080] Wherein, (x″, y″) is the coordinate after translation; Δx and Δy are the translation amounts in the x and y directions.
[0081] The image scaling formula is:
[0082]
[0083] Wherein, (x″′, y″′) is the scaled coordinate; s is the scaling factor.
[0084] S23, constructing a U-Net model and inputting a mask layer image for training;
[0085] U-Net is a classic convolutional neural network architecture, which is widely used in image segmentation tasks. In this embodiment, for PCB mask layer defect detection, the input PCB image can be segmented into defective areas and normal areas, so as to identify defects in the mask layer, such as bubbles, wrinkles, etc.
[0086] U-Net is a symmetrical encoder-decoder network structure, which includes an encoder part, a bridge part, and a decoder part;
[0087] 1) The encoding part includes:
[0088] Convolutional layer: extracts image features through a series of convolution operations;
[0089] Activation function: Use ReLU (Rectified Linear Unit) activation function;
[0090] Pooling layer: Use Max Pooling to reduce the spatial dimension of the image.
[0091] 2) The bridge part includes a convolutional layer to process high-dimensional features and extract complex features;
[0092] 3) The decoding part includes:
[0093] Upsampling layer: restores the spatial resolution of the image through transposed convolution or upsampling operations.
[0094] Convolutional layer: Connects to the image features of the encoded part, fuses low-level features and high-level features through convolution operations, and generates a predicted segmentation map of the mask layer.
[0095] The loss function of the U-Net model is:
[0096]
[0097] where y i is the true label, p i is the predicted probability.
[0098] The steps to train the U-Net model are:
[0099] S231: dividing the preprocessed mask layer image into a training set and a validation set in a ratio of 8:2;
[0100] S232, input the training set data into the U-Net model for training, and update the weights through forward propagation and back propagation;
[0101] S233, at the end of each epoch, the validation set is used to evaluate the performance of the model, and the hyperparameters are adjusted based on the performance evaluated on the validation set.
[0102] S23, uses the trained U-Net model to analyze the image and identify defects in the mask layer, such as bubbles and wrinkles.
[0103] S3, gold plating is performed on the PCB that has been tested and has no mask layer defects;
[0104] The specific steps include:
[0105] S31, performing acid etching on the mask layer to remove surface dirt;
[0106] The acid etching solution used in the acid etching treatment is a mixture of one or more of hydrochloric acid, sulfuric acid or hydrofluoric acid;
[0107] The specific process of acid etching is as follows:
[0108] S311, immerse the PCB completely in the acid etching solution; the immersion time is between 1 and 10 minutes;
[0109] S312, during the immersion process, stirring the acid etching solution can improve the uniformity of the solution and the etching efficiency; a bubble stirring or mechanical stirring device can be used;
[0110] S313, regularly check the surface of the PCB board and observe the progress of acid etching to ensure that surface dirt and oxides are effectively removed;
[0111] S314, rinse the PCB board with deionized water or distilled water to remove the residual acid etching solution; the rinsing time is usually between 1-5 minutes.
[0112] S315, air dry the cleaned PCB board or use hot air drying equipment to dry it to ensure that the surface is dry and free of water marks.
[0113] S32, gold-plating the PCB using electroplating gold process;
[0114] The electroplating solution contains gold salt, and the plating current and voltage need to be precisely controlled. The current density is generally 0.1-0.5A / dm 2 , the temperature is controlled between 45-55℃.
[0115] S33, cleaning the gold-plated layer to remove residual chemicals and drying it to obtain a gold-plated PCB board.
[0116] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0117] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.
Claims
1. A process for manufacturing a gold-plated PCB with anti-seepage plating, characterized in that: The following steps are involved: S1, cleaning the PCB and coating a photosensitive masking agent on the surface of the PCB to form a masking layer; S2, collects the mask layer image and uses the U-Net model to analyze the mask layer image to achieve defect recognition and detection; S3, gold-plating the PCB that has been detected to have no mask layer defects to obtain a gold-plated PCB.
2. The manufacturing process of a PCB gold-plated plate with anti-seepage plating according to claim 1, characterized in that: The photosensitive masking agent is: polyimide, photosensitive resin, phenolic resin, photosensitive polyurethane or photosensitive silicone rubber.
3. The process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 1, characterized in that: A spin coater is used to coat the photosensitive mask agent, the spin coating speed is 500-600r / min, and the mask layer thickness is 3-8um.
4. The manufacturing process of a PCB gold-plated plate with anti-seepage plating according to claim 1, characterized in that: The U-Net model includes: an encoding part, a bridging part and a decoding part; The encoding part includes: Convolutional layer: extracts image features through a series of convolution operations; Activation function: Use ReLU activation function; Pooling layer: Use maximum pooling to reduce the spatial dimension of the image; The bridge part includes a convolution layer to process high-dimensional features and extract complex features; The decoding part includes: Upsampling layer: restores the spatial resolution of the image through transposed convolution or upsampling operations; Convolutional layer: Connects to the image features of the encoded part, fuses low-level features and high-level features through convolution operations, and generates a predicted segmentation map of the mask layer.
5. A process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 4, characterized in that: The steps of using the U-Net model to detect defects in the mask layer image include: S21, collecting an image of the mask layer using a CCD camera or a microscope camera; S22, preprocessing the collected mask layer image; S23, input the preprocessed image into the U-Net model for training; S24, using the trained U-Net model to segment the mask layer image and output defect recognition results.
6. A process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 5, characterized in that: The preprocessing of the mask layer image includes: normalization processing and data enhancement processing; The normalization formula is: Where I is the original pixel value; I min and I max are the minimum and maximum pixel values in the image respectively; I norm is the normalized pixel value in the range of [-1, 1].
7. A process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 6, characterized in that: The data enhancement processing includes: rotation, translation and scaling of the image; The formula for image rotation is: Where (x, y) is the original coordinate, (x′, y′) is the rotated coordinate; θ is the rotation angle; The image translation formula is: Where (x″, y″) is the coordinate after translation; Δx and Δy are the translation amounts in the x and y directions; The image scaling formula is: Wherein, (x″′, y″′) is the scaled coordinate; s is the scaling factor.
8. The process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 4, characterized in that: The loss function of the U-Net model is: Among them, y i is the true label, p i is the predicted probability.
9. The process for manufacturing a gold-plated PCB with anti-seepage plating according to claim 1, characterized in that: The steps for PCB gold plating include: S31, performing acid etching on the mask layer; S32, gold-plating the PCB using electroplating gold process; S33, cleaning the gold-plated layer, and drying it to obtain a gold-plated PCB.
10. A system for manufacturing PCB gold-plated plates with anti-seepage plating, characterized in that: include: Mask layer coating module: cleans the PCB and applies a photosensitive masking agent on the PCB surface to form a mask layer; Image analysis and detection module: collects mask layer images and uses the U-Net model to analyze the mask layer images to achieve defect recognition and detection; And, a gold-plating module: gold-plating is performed on the PCB that has been detected to have no mask layer defects to obtain a gold-plated PCB board.
Citation Information
Cited By
PCB (Printed Circuit Board) defect identification method used in automatic PCB sorting process
CN119579584A
