A printed circuit board back-drilling drilling depth control method
By using an improved density peak clustering algorithm and a two-dimensional weighted compensation model, the problem of inaccurate depth caused by material properties and thickness fluctuations during back drilling of printed circuit boards was solved, achieving precise back drilling control and high-yield production.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- FOSHAN SHUNDE JUNDA ELECTRONIC CO LTD
- Filing Date
- 2026-03-20
- Publication Date
- 2026-06-23
Smart Images
Figure CN122269575A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of printed circuit board processing technology, and specifically relates to a method for controlling the drilling depth of back drilling in printed circuit boards. Background Technology
[0002] As printed circuit boards (PCBs) evolve towards higher speeds and frequencies, the precision requirements for back drilling are becoming increasingly stringent. The purpose of back drilling is to remove redundant portions of vias (PTHs) that are not used for signal transmission, known as stubs, to reduce signal reflection and loss and ensure impedance continuity. Accurately determining the back drilling depth is crucial to ensuring that the stub length meets requirements.
[0003] Currently, the mainstream method for determining back-drilling depth in the industry is the preset depth method. This method pre-sets the back-drilling depth based on the theoretical thickness of each dielectric and copper layer in the PCB design stack-up. However, in actual production, this method faces many challenges: First, factors such as uneven resin flow in the prepreg (PP) and core thickness tolerances during PCB lamination cause deviations between the actual and theoretical thicknesses of each layer. Even within the same batch of PCBs, thickness variations exist, directly impacting the accuracy of back-drilling depth. Second, significant differences in material properties exist across different areas of the PCB. For example, the resin filling density differs between the edge and center areas, the fiberglass weaving direction leads to dielectric anisotropy, and uneven copper foil residual copper distribution affects the uniformity of thickness after lamination. Existing methods using a uniform back-drilling depth across the entire board cannot accommodate these regional differences. Furthermore, drill wear accumulates with the number of drilled holes, causing deviations between the actual and set back-drilling depths. The heat accumulation effect alters material properties, further affecting back-drilling accuracy. Traditional preset depth methods, using fixed compensation values, struggle to address these dynamic changes. Summary of the Invention
[0004] To address the aforementioned problems in the prior art, this invention provides a method for controlling the back-drilling depth of printed circuit boards, which solves the problem of accurately determining the back-drilling depth of each area due to differences in material properties and thickness fluctuations in different regions of the PCB, so as to ensure the precise controllability of the residual pile length.
[0005] The objective of this invention can be achieved through the following technical solution: a method for controlling the drilling depth of back-drilling holes in printed circuit boards, the method comprising the following steps: S1: In the PCB design file to be processed, extract the vertical theoretical thickness from the back-drilling start layer to the back-drilling end target signal layer, the PCB board thickness tolerance parameters, the preset residual length, and the material property parameters of each dielectric layer and copper foil layer between the back-drilling start layer and the target signal layer. S2: Based on the stack-up layout, material distribution characteristics, and copper foil ratio distribution in the PCB design file, the PCB board is divided into multiple anisotropic sub-back-drill areas according to the preset area division rules. S3: Perform multi-point actual measurement of plate thickness in each sub-back-drilling area to obtain plate thickness data sequence and calculate the area thickness characteristic value; S4: Based on the material property parameters and thickness characteristic values of each sub-back drill region, calculate the thickness compensation coefficient of each sub-back drill region using the set weighted model. S5: Calculate the back-drilling depth of each sub-back-drilling region based on the thickness characteristic value, thickness compensation coefficient, and preset residual pile length of the sub-back-drilling region; S6: Establish the mapping relationship between the PCB board design coordinates and the sub-back drilling area, match and assign the back drilling depth to all back drilling target holes in the sub-back drilling area, and control the CNC drilling machine to perform back drilling processing according to the back drilling depth of the sub-back drilling area corresponding to each hole.
[0006] Preferably, in S6, during the back drilling process, the heat accumulation parameters during the continuous drilling process are also monitored in real time. The heat accumulation parameters include the number of continuous drillings, the average distance between adjacent holes, and the instantaneous drilling load. The heat accumulation compensation amount is calculated based on the heat accumulation parameters. The calculated heat accumulation compensation amount is then continuously calculated and compensated for in real time for the back drilling depth, and the back drilling processing program of the CNC drilling machine in S6 is updated.
[0007] Preferably, in S6, the formula for calculating the heat product compensation is: ; In the formula, This is the amount of heat product compensation. As the baseline thermal compensation amount, This refers to the number of consecutive boreholes. This is a reference value for the maximum number of consecutive boreholes. This is the continuous drilling effect coefficient. The power exponent for continuous borehole drilling. This represents the average distance between adjacent holes. The attenuation coefficient of the hole spacing is... This represents the instantaneous drilling load during the current drilling process. This is the reference drilling load for the drilling rig.
[0008] Preferably, in step S2, dividing the PCB board into multiple anisotropic sub-back-drilling regions specifically includes the following steps: S21: Calculate the local residual copper ratio of each layer at the location point based on the stack-up layout, material distribution characteristics and copper foil ratio distribution in the PCB design file; S22: The equivalent residual copper rate at each location point is calculated using a thickness-weighted average model; S23: Extract the material property feature parameters of each dielectric layer to construct a material property feature vector. The material property feature parameters include the conductivity parameters of the dielectric layer, the fiberglass cloth weaving angle, the resin content percentage, and the dielectric layer thickness. S24: The equivalent residual copper rate and material property feature vector of the coordinate position point are fused to obtain a comprehensive feature vector; S25: The improved density peak clustering algorithm is used to perform cluster analysis on the comprehensive feature vector, dividing the PCB board into multiple anisotropic sub-back-drill regions.
[0009] Preferably, in S22, the formula for calculating the equivalent residual copper rate is: ; In the formula, Location point The equivalent residual copper rate, Let be the theoretical design thickness of the i-th layer, and n be the total number of layers from the back-drilling start layer to the target signal layer. Let be the local residual copper ratio of the i-th layer at the location point.
[0010] Preferably, S25 specifically includes the following sub-steps: S251: Calculate the weighted Euclidean distance between any two locations, and calculate the local density and relative distance of each location based on the weighted Euclidean distance; S252: Select the location points that simultaneously satisfy the local density greater than the local density threshold and the relative distance greater than the relative distance threshold as cluster centers. The two thresholds are adaptively determined by the inflection point method of the cluster decision graph. S233: Assign non-cluster center locations to the regions containing the nearest cluster centers to divide the initial sub-back-drill regions; S254: Morphological closure operation is used to perform boundary smoothing optimization on the initial sub-backdrill region to determine each sub-backdrill region; S255: Output the results of each sub-drilling region division, including the coordinate boundaries and region identifiers of each sub-region.
[0011] Preferably, in S251, the formula for calculating the weighted Euclidean distance is: ; In the formula, Let be the weighted Euclidean distance between point p and point q. , These are the normalized values of the k-th dimension of the comprehensive feature vector between positions p and q. denoted as the weight coefficient of the k-th dimension feature.
[0012] Preferably, in S3, the region thickness characteristic values include the region thickness standard deviation, region thickness skewness, and region thickness kurtosis.
[0013] Preferably, in step S4, the thickness compensation coefficient for each sub-back drill region is calculated using a set weighted model, including the following steps: S41: Calculate the material property compensation factor for the sub-back drill region based on the eigenvalues in the normalized material property eigenvector. S42: Calculate the thickness feature compensation factor of the sub-back drill region based on the normalized thickness feature values of the three regions; S43: By using a preset two-dimensional weighted model, the material property compensation factor and the thickness feature compensation factor are fused together to calculate the thickness compensation coefficient of each sub-back drill region.
[0014] Preferably, S5 specifically includes the following sub-steps: S51: For each sub-back-drilling area, based on the average thickness of the area and combined with the stacking theory parameters in the PCB design file, calculate the actual effective thickness from the starting layer to the target signal layer. S52: Based on the actual effective thickness of the sub-region, the thickness compensation coefficient, and the preset residual pile length, the back-drilling depth of each sub-back-drilling region is calculated; S53: Perform back-drilling depth safety threshold verification to determine the back-drilling depth of each sub-region.
[0015] The beneficial effects of this invention are as follows: This invention utilizes an improved density peak clustering algorithm that fuses equivalent residual copper ratio and material property feature vectors to adaptively and accurately divide the anisotropic sub-drilling regions within the PCB board. Simultaneously, it constructs a two-dimensional weighted compensation model that integrates inherent material processing characteristics and measured thickness statistical features, replacing the traditional single fixed empirical compensation. Furthermore, in subsequent back-to-back drilling processes, a real-time compensation mechanism for continuous drilling heat accumulation is implemented. Through differentiated and precise compensation across regions, it solves the problem of distorted back-drilling depth reference caused by uneven thickness and material property differences within the PCB during lamination. This improves the accuracy of back-drilling residual length control while reducing the risk of scrapping the target signal layer and increasing the mass production yield in PCB back-drilling processing. Attached Figure Description
[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.
[0017] Figure 1 This is a schematic diagram of the steps of the control method of the present invention; Figure 2 This is a schematic diagram of the sub-step S2 in the control method of the present invention; Figure 3This is a schematic diagram of sub-step S5 in the control method of the present invention. Detailed Implementation
[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.
[0019] Please see Figures 1-3 This embodiment provides a method for controlling the drilling depth of back-drilling holes in printed circuit boards, the method including the following steps: S1: Obtain the PCB design file to be processed, and extract the vertical theoretical thickness from the back-drilling start layer to the back-drilling end target signal layer, the PCB board thickness tolerance parameters, the preset residual length, and the material property parameters of each dielectric layer and copper foil layer between the back-drilling start layer and the target signal layer. S2: Based on the stack-up layout, material distribution characteristics, and copper foil ratio distribution in the PCB design file, the PCB board is divided into multiple anisotropic sub-back-drilling areas according to the preset area division rules. S2 specifically includes the following sub-steps: S21: Based on the stack-up layout, material distribution characteristics, and copper foil ratio distribution in the PCB design file, calculate the local residual copper rate of each layer at the location point. Specifically: Divide the entire plane of the PCB to be processed into a×b grid cells according to a preset resolution. Each grid cell corresponds to a unique planar coordinate location point. The minimum size of the grid cell is adapted to the minimum routing precision of the PCB design file, for each coordinate point. The calculation scope covers all layers from the back-drilling start layer to the back-drilling end target signal layer in the PCB design file. The local residual copper rate is calculated layer by layer within the corresponding grid point. The local residual copper rate is the ratio of the effective copper foil area of the corresponding layer within that grid cell to the total area of the grid cell. The calculation formula is as follows: ; In the formula, The local residual copper ratio of the i-th layer at coordinate location point. This represents the effective area of the i-th layer of copper foil within the grid corresponding to this location point. The total area of a single grid cell is used to quantify the copper foil distribution density at the corresponding location. The significant differences in the pressing shrinkage rate and drilling load characteristics between the copper foil and the dielectric material are the core causes of uneven pressing thickness and back-drilling load fluctuations within the PCB board, providing core characteristics for the copper foil distribution dimension in subsequent area division.
[0020] S22: Calculate the equivalent residual copper rate at each location point, specifically for each location point. A thickness-weighted average model is used to integrate the local residual copper rates of multiple layers in the vertical stacking direction into a single-value equivalent residual copper rate at that location. This eliminates the dimensional differences in the multilayer stacking and achieves vertical comprehensive quantification of the copper foil distribution characteristics. The calculation formula for the thickness-weighted average model is as follows: ; In the formula, Location point The equivalent residual copper rate, Let be the theoretical design thickness of the i-th layer, and n be the total number of layers from the back-drilling start layer to the target signal layer. The comprehensive copper foil distribution ratio, weighted by the thickness of the entire back-drilling path, is used to quantify the lamination thickness characteristics and drilling load characteristics at this location. The thicker the stack, the greater the impact on the overall characteristics of the back-drilling path, and the higher the corresponding weight. This ensures that the equivalent residual copper rate can truly reflect the comprehensive copper foil distribution characteristics on the back-drilling path at this location, and avoids the interference of abnormal copper foil distribution of thin stacks on feature extraction.
[0021] S23: For each location point, construct a material property feature vector for each medium layer between the back-drilling start layer and the target signal layer, quantify the differences in material processing characteristics at that location point, and the expression for the material property feature vector is: ; in, These are the dielectric layer conductivity parameters, fiberglass cloth weave angle, resin content percentage, and dielectric layer thickness, respectively. The dielectric layer conductivity parameters are specifically the weighted average of the bulk resistivity and surface resistivity of the dielectric layer, used to characterize the electrical properties of the dielectric layer and the electrical signal feedback characteristics during the drilling process. The fiberglass cloth weave angle of the dielectric layer is used to characterize the anisotropic structural characteristics of the dielectric layer, directly affecting the Z-axis thermal expansion coefficient and compression shrinkage rate of the dielectric layer. The resin content percentage of the dielectric layer is used to characterize the resin-to-fiberglass cloth ratio of the dielectric layer, directly affecting the drilling hardness, thermal conductivity, and compression deformation characteristics of the dielectric layer. h is the weighted average thickness of the dielectric layer, calculated with the thickness of each dielectric layer as the weight.
[0022] Material property parameters are extracted from the stacked material table in the PCB design file and weighted by the dielectric layer thickness to ensure that the material property feature vector corresponds to each coordinate point.
[0023] S24: First, normalize the equivalent residual copper rate and material property feature vector of the coordinate position point, and then splice and fuse the normalized data to obtain a comprehensive feature vector.
[0024] S25: The improved Density Peak Clustering (DCP) algorithm is used to perform cluster analysis on the comprehensive feature vector, dividing the PCB board into multiple anisotropic sub-drilling regions. The DCP algorithm specifically includes the following execution process: S251: Calculate the weighted Euclidean distance between any two points within the global PCB area. Based on the weighted Euclidean distance, calculate the local density and relative distance of each point. The formula for calculating the weighted Euclidean distance is: ; In the formula, Let be the weighted Euclidean distance between point p and point q. , These are the normalized values of the k-th dimension of the comprehensive feature vector between positions p and q. The weighting coefficients for the k-th dimension feature are, in this embodiment, the equivalent residual copper rate weight. =0.4, average thickness weight of dielectric layer =0.2, and the sum of the weights of the remaining features is 0.4. The weight coefficients can be adjusted through PCB back-drilling process test calibration.
[0025] Local density, used to represent the density of points around a given location, is calculated using a Gaussian kernel function, which sums the contributions of all other points to that location. The formula is as follows: ; In the formula, The cutoff distance is determined by the percentile value of the weighted Euclidean distance of all locations across the PCB. In this embodiment, the 2% percentile value is used to ensure the anti-interference capability of the local density calculation.
[0026] The relative distance is the minimum weighted Euclidean distance between a location point and a location point with a relatively higher local density. It is used to measure the degree of separation between the point and the high-density region. For the location point with the highest local density in the entire region, the relative distance is the largest weighted Euclidean distance in the entire region.
[0027] S252: Construct a clustering decision graph with local density as the horizontal axis and relative distance as the vertical axis. Use local density threshold and relative distance threshold as cluster center determination indicators. Select the location points that simultaneously satisfy the local density greater than the local density threshold and the relative distance greater than the relative distance threshold as cluster centers. The two thresholds are adaptively determined by the inflection point method of the clustering decision graph.
[0028] S233: All non-cluster center locations are assigned to the cluster center of the nearest location with a local density greater than that location, according to the principle of minimizing weighted Euclidean distance. Each location belongs to only one cluster center, forming an initial sub-drilling region corresponding to the number of cluster centers. This ensures that the difference in the comprehensive feature vector within each initial sub-region is minimized, and the difference in the comprehensive feature vector between intervals is maximized.
[0029] S254: The boundary of the initial sub-drilling region is smoothed and optimized. Morphological closure operation is used to process the region boundary, fill the small holes in the boundary, and eliminate the jagged irregular boundary to obtain a continuous and smooth region outline. In this embodiment, the morphological closure operation uses a 3×3 rectangular structural element to adapt to the grid cell size, ensuring that the core coverage of the sub-region is not changed after the boundary smoothing process, and the boundary line does not cross any back-drilling target hole position, avoiding the problem of a single back-drilling hole spanning two or more sub-regions.
[0030] S255: Sub-back-drilling area division results for the PCB board, including the unique area identifier of each sub-back-drilling area, the planar coordinate boundary (minimum bounding rectangle and contour coordinate point set), the coordinates of all back-drilling target holes contained in the area, and the average comprehensive feature vector of the area.
[0031] Step S2, the sub-back-drilling region division, quantifies the spatial anisotropy within the PCB board through multi-dimensional fusion of equivalent residual copper rate and material property characteristics. The back-drilling processing characteristics within the divided sub-regions are highly uniform, avoiding errors caused by uniform compensation across the entire board. An improved density peak clustering algorithm is adopted, eliminating the need to pre-set the number of sub-regions. It can adaptively complete the division based on the actual characteristics of the PCB board, adapting to PCB boards with different stack-ups, materials, and layouts, and has strong versatility. At the same time, it avoids errors caused by manually setting the number of regions. The sub-regions after boundary smoothing can be directly connected to the processing program of CNC drilling machines, adapting to PCB mass production processing technology.
[0032] S3: For each sub-back-drilled area, perform multi-point actual thickness measurements to obtain a thickness data sequence and calculate the area thickness characteristic values. The area thickness characteristic values include the area average thickness, area minimum thickness, area maximum thickness, area thickness standard deviation, area thickness skewness, and area thickness kurtosis. The specific process includes the following steps: For each sub-back-drilled area, based on the area's physical dimensions and back-drill hole distribution density, an equidistant grid layout rule was adopted. Multi-point plate thickness measurements were conducted using a contact laser thickness gauge. The resulting raw plate thickness data sequence was preprocessed to eliminate measurement errors and outlier interference. The layout principle was as follows: When the area of the sub-region is ≤100mm 2 At that time, a total of 9 thickness measurement points were set up in a 3×3 pattern to cover the four corners, the midpoints of the four sides, and the center of the sub-region; When the area of the sub-region is greater than 100mm 2 At the same time, 25 thickness measurement points in a 5×5 grid were set up to evenly cover the entire sub-region, with a focus on increasing the density of thickness measurement points in areas with dense back boreholes. Furthermore, all thickness measurement points must avoid PCB traces, pads, vias, and board edge positioning holes to ensure that the thickness measurement data only reflects the true thickness of the PCB substrate and is free from external structural interference.
[0033] Based on the preprocessed effective plate thickness data sequence, the regional thickness characteristic value of the sub-back-drilling region is calculated: The average thickness of the area is the arithmetic mean of the thickness of the plate at all valid thickness measurement points within the sub-back drilling area; The minimum thickness and maximum thickness of the region are the minimum and maximum values in the effective board thickness data sequence of the sub-back drilling region, respectively. The difference between the two is the region thickness range, which is used to characterize the thickness fluctuation limit range within the sub-region. This provides a boundary basis for setting the safety threshold of back drilling depth and avoids drilling through the target signal layer, which would lead to PCB scrapping. The standard deviation of the area thickness is used to quantify the dispersion of the plate thickness within the sub-drilling area. The specific calculation formula is as follows: ; In the formula, Let n be the standard deviation of the thickness in the region, and n represent the total number of valid thickness measurement points. The average thickness of the region. The measured plate thickness is the value of the i-th effective thickness measurement point within the sub-back drilling area. Regional thickness skewness, used to characterize the asymmetry of the thickness data distribution in the sub-back drilling region, is calculated using the following formula: ; A positive regional thickness skewness indicates that the plate thickness data is shifted in a direction greater than the average thickness; a negative regional thickness skewness indicates that the thickness is shifted in a direction less than the average thickness. This is used to correct the compensation reference skewness caused by asymmetric thickness distribution. Region thickness kurtosis, used to represent the steepness of the thickness data distribution in the sub-back drilling region, is calculated using the following formula: ; The larger the kurtosis of the regional thickness, the more concentrated the plate thickness data is around the average thickness, the better the thickness uniformity of the sub-region, and the higher the stability of the compensation coefficient.
[0034] The calculated average thickness, minimum thickness, maximum thickness, standard deviation, skewness, and kurtosis of the region are normalized respectively.
[0035] S4: Based on the material property feature vectors and thickness feature values of each sub-back-drilling region, the thickness compensation coefficient of each sub-back-drilling region is calculated using a set weighted model. By integrating the inherent processing characteristics of the material in each sub-back-drilling region with the measured thickness distribution characteristics, a two-dimensional weighted compensation model is constructed. This addresses the shortcomings of existing technologies, such as the inability of fixed empirical compensation values to adapt to differences in regional material and thickness, insufficient compensation accuracy, and inability to handle anisotropy within the board. The specific implementation process is as follows: S41: Based on the eigenvalues in the normalized material property eigenvector, calculate the material property compensation factor for the sub-back drill region. This factor is used to correct the deviation between the set drilling depth and the actual drilling depth caused by the inherent properties of the material (hardness, thermal expansion, drilling resistance). The calculation formula is as follows: ; In the formula, This is the material property compensation factor for the i-th sub-back drill region, with a value between 0 and 1. - These are the weighting coefficients for each dimension of each material property. In this embodiment, =0.2, =0.25, =0.35, =0.2, the sum of the weighting coefficients is 1, which can be adjusted through PCB back drilling process test calibration. The percentage of resin content can be set to the maximum weight value.
[0036] S42: Based on the normalized thickness characteristic values of the three regions, calculate the thickness characteristic compensation factor for the sub-drilling region. This factor is used to correct the back-drilling depth reference deviation caused by the fluctuation and uneven distribution of the measured thickness in the sub-region. The calculation formula is as follows: ; In the formula, This is the thickness characteristic compensation factor, ranging from 0 to 1. A larger value indicates poorer thickness uniformity and more significant thickness distribution deviation in the sub-region, requiring a larger amount of drilling depth compensation correction. - These are the weighting coefficients for each dimension of each material property. In this embodiment, =0.5, =0.3, =0.2, the sum of the weighting coefficients is 1, which can be adjusted through process test calibration.
[0037] S43: By using a pre-defined two-dimensional weighted model, the material property compensation factor and the thickness feature compensation factor are fused to calculate the thickness compensation coefficient for each sub-back drill region. The calculation formula is as follows: ; In the formula, The reference compensation coefficient is the basic compensation value under standard operating conditions with no material deviation and no thickness fluctuation. It is determined through process calibration tests on a standard PCB board. In this embodiment... =0.98, The total weight of the material property compensation factor. The total weight of the thickness feature compensation factor satisfies + =1; In this embodiment, for high-speed, high-frequency PCB boards, the material properties have a more significant impact on the drilling depth, so the setting is... =0.6, =0.4; For ordinary FR-4 material PCB boards, set =0.4, =0.6, which can be flexibly adjusted according to the type of board and process requirements.
[0038] The final thickness compensation coefficient is a correction parameter for calculating the back drilling depth of the sub-back drilling area. Each sub-area corresponds to a unique exclusive compensation coefficient, realizing precise differential compensation for different areas and fully adapting to the material characteristics and thickness distribution characteristics of different sub-areas.
[0039] S5: Calculate the back-drilling depth of each sub-back-drilling region based on the thickness characteristic value, thickness compensation coefficient, and preset residual pile length of the sub-back-drilling region. This includes the following steps: S51: For each sub-drilling area, based on the average thickness of the area obtained in step S3, and combined with the stacking theory parameters in the PCB design file, the actual effective thickness from the initial layer to the target signal layer is calculated in reverse. This eliminates the influence of overall board thickness shrinkage or expansion on the back-drilling path thickness during lamination. The formula for calculating the effective thickness over time is: ; In the formula, The average thickness of the sub-back drill area calculated in S3 (measured value). In PCB design files, this refers to the theoretical vertical thickness from the start layer of the back-drilling process to the target signal layer at the end of the back-drilling process. This refers to the theoretical total thickness of the entire board in the PCB design file.
[0040] S52: Based on the actual effective thickness of the sub-region, the thickness compensation coefficient, and the preset residual pile length, the back-drilling depth of the sub-back-drilling region is calculated. The calculation formula is: ; In the formula, K is the thickness compensation coefficient of the sub-back drill region calculated by S4. The preset residual stud length in the PCB design documents and process specifications is, in this embodiment, ≤0.075mm (3mil) for high-speed and high-frequency PCBs above 25G, and ≤0.125mm (5mil) for ordinary consumer-grade PCBs. This can be flexibly adjusted according to customer signal rate requirements and industry standards. After correcting drilling deviations with a thickness compensation coefficient, it is ensured that the drill bit stops just above the target signal layer after drilling, and the remaining residual stud length fully meets the design requirements, while minimizing the risk of drilling through the target signal layer.
[0041] S53: Perform a safety threshold verification for the back drilling depth. Based on the calculated back drilling depth D of the sub-region and the minimum thickness of the region obtained in S3, perform a safety threshold verification to avoid drill-through failure caused by extreme thickness fluctuations. The verification formula is as follows: ; In the formula, represents the minimum thickness of the region calculated in S3, and the minimum effective thickness of the back-drilling path calculated according to the above effective thickness calculation formula. To provide a safety margin for the budget based on the PCB design requirements, if the calculated back-drilling depth exceeds the upper limit of the safety threshold, the upper limit of the safety threshold will be used as the final back-drilling depth for that sub-region. Under the premise that the length of the residual pile is controllable, the yield of back-drilling processing will be maximized.
[0042] S54: The verified back-drilling depth is used as the final back-drilling depth of the sub-back-drilling region. It is bound and stored with the sub-region's unique region identifier, coordinate boundary, and list of included back-drilling target holes, providing accurate parameter basis for the hole matching assignment in the subsequent S6.
[0043] S6: Establish the mapping relationship between the PCB board design coordinates and the sub-back-drilling area, assign the back-drilling depth to all back-drilling target holes in the sub-back-drilling area, and control the CNC drilling machine to perform back-drilling processing according to the back-drilling depth of the sub-back-drilling area corresponding to each hole. Specifically, this includes: Obtain the design coordinates of all back-drill target holes in the PCB design file. Based on the boundary coordinates and coordinate identifiers of the sub-back-drill areas divided in step S2, determine the sub-back-drill area to which each back-drill target hole belongs and determine the mapping relationship between the back-drill target hole and the sub-back-drill area. The calculated back-drilling depth of the sub-back-drilling region is assigned to all back-drilling target holes in the corresponding sub-back-drilling region. Generate a back-drilling machining program containing the coordinates of each target hole and the corresponding back-drilling depth. Use the back-drilling machining program to control the CNC drilling machine to perform back-drilling machining according to the back-drilling depth corresponding to each target hole.
[0044] During continuous drilling, the drill bit generates heat through high-speed friction with the PCB material. Due to the low thermal conductivity of the PCB material (fiberglass cloth, resin), the heat cannot dissipate quickly enough, causing the drill bit temperature to rise. Once the resin reaches its glass transition temperature, its mechanical and dielectric properties undergo abrupt changes. The PCB material expands due to heat, leading to variations in the actual layer thickness. The increased temperature also causes changes in contact resistance, resulting in attenuation and waveform distortion of the acquired electrical signal amplitude. Consequently, the signal characteristics of holes drilled earlier and later on the same PCB become inconsistent. Without intervention, the error in determining the back-drilling depth gradually increases with the processing progress. While existing technologies include methods to cool the drill bit, this increases processing time and, while improving quality, hinders speed increases.
[0045] Therefore, further, during the back drilling process, the heat accumulation parameters during continuous drilling are monitored in real time. These parameters include the number of consecutive holes drilled, the average distance between adjacent holes, and the instantaneous drilling load. The heat accumulation compensation is calculated, and the calculated heat accumulation compensation is continuously calculated and compensated for in real time to update the back drilling program. The formula for calculating the heat accumulation compensation is as follows: ; In the formula, This is the amount of heat product compensation. This is the baseline thermal compensation amount, the basic compensation value under standard operating conditions (single hole drilling, reference hole spacing, reference load). The current number represents the number of consecutive holes drilled, while the previous number represents the number of holes drilled consecutively. This is a reference value for the maximum number of consecutive holes, set according to the equipment's heat dissipation capacity, such as 100 holes. This is the continuous drilling effect coefficient. The power exponent for continuous borehole drilling. This represents the average distance between adjacent holes, specifically the average distance between the current hole and the nearest few drilled holes. For reference hole spacing, and The dimensions are consistent. The attenuation coefficient of the hole spacing is... The power exponent. The instantaneous drilling load during the current drilling process is calculated from the drilling rig spindle current, power value, or equivalent current load. This refers to the reference drilling load for the drilling rig, specifically the rated reference load value of the drilling rig preset in the process. and The dimensions are consistent.
[0046] The heat product compensation is calculated using a three-factor product structure. Based on the principle of superposition of temperature fields from multiple heat sources in heat conduction theory, the continuous drilling number factor adopts a power law form, reflecting the temperature accumulation law under the continuous action of heat sources; the hole spacing factor adopts an exponential decay form, which conforms to the solution of the point heat source temperature field decaying exponentially with distance; the drilling load factor adopts a power law form, reflecting the nonlinear relationship between heat generation power and load, which is used to solve the problem of heat-induced signal drift caused by the heat accumulation effect during drilling, and to ensure the stability and consistency of back drilling depth determination throughout the entire machining process.
[0047] This application addresses the technical problems of existing PCB back-drilling processes, where a uniform preset depth and fixed experience-based compensation scheme for the entire board cannot adapt to the anisotropic differences in material distribution, copper foil ratio, and lamination thickness within the board. Furthermore, it cannot reduce the degradation of back-drilling accuracy caused by continuous drilling heat accumulation without cooling the drill bit. This application utilizes an improved density peak clustering algorithm that fuses equivalent residual copper rate and material property feature vectors to adaptively and accurately divide anisotropic sub-back-drilling regions within the PCB board. Simultaneously, it constructs a two-dimensional weighted compensation model that integrates inherent material processing characteristics and measured thickness statistical features, replacing the traditional single fixed experience-based compensation. This further enables a real-time compensation mechanism for continuous drilling heat accumulation in subsequent back-drilling processes, achieving comprehensive and precise control of back-drilling depth from static benchmark correction to dynamic processing deviation compensation.
[0048] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. A method for controlling the drilling depth of back-drilling holes in a printed circuit board, characterized in that: The method includes the following steps: S1: In the PCB design file to be processed, extract the vertical theoretical thickness from the back-drilling start layer to the back-drilling end target signal layer, the PCB board thickness tolerance parameters, the preset residual length, and the material property parameters of each dielectric layer and copper foil layer between the back-drilling start layer and the target signal layer. S2: Based on the stack-up layout, material distribution characteristics, and copper foil ratio distribution in the PCB design file, the PCB board is divided into multiple anisotropic sub-back-drill areas according to the preset area division rules. S3: Perform multi-point actual measurement of plate thickness in each sub-back-drilling area to obtain plate thickness data sequence and calculate the area thickness characteristic value; S4: Based on the material property parameters and thickness characteristic values of each sub-back drill region, calculate the thickness compensation coefficient of each sub-back drill region using the set weighted model. S5: Calculate the back-drilling depth of each sub-back-drilling region based on the thickness characteristic value, thickness compensation coefficient, and preset residual pile length of the sub-back-drilling region; S6 Establish a mapping relationship between the PCB board design coordinates and the sub-back-drilling area, match and assign the back-drilling depth to all back-drilling target holes in the sub-back-drilling area, and control the CNC drilling machine to perform back-drilling processing according to the back-drilling depth of the sub-back-drilling area corresponding to each hole.
2. The method for controlling the drilling depth of back-drilling holes in a printed circuit board according to claim 1, characterized in that: In S6, during the back drilling process, the heat accumulation parameters during the continuous drilling process are also monitored in real time. The heat accumulation parameters include the number of consecutive holes, the average distance between adjacent holes, and the instantaneous drilling load. The heat accumulation compensation amount is calculated based on the heat accumulation parameters. The calculated heat accumulation compensation amount is then continuously calculated and compensated for in real time for the back drilling depth, and the back drilling processing program of the CNC drilling machine in S6 is updated.
3. The method for controlling the drilling depth of back-drilling holes in a printed circuit board according to claim 1, characterized in that: In S6, the formula for calculating the heat product compensation is: ； In the formula, This is the amount of heat product compensation. As the baseline thermal compensation amount, This refers to the number of consecutive boreholes. This is a reference value for the maximum number of consecutive boreholes. This is the continuous drilling effect coefficient. The power exponent for continuous borehole drilling. This represents the average distance between adjacent holes. The attenuation coefficient of the hole spacing is... This represents the instantaneous drilling load during the current drilling process. This is the reference drilling load for the drilling rig.
4. The method for controlling the drilling depth of back-drilling holes in a printed circuit board according to claim 1, characterized in that: In S2, dividing the PCB board into multiple anisotropic sub-back-drilling regions specifically includes the following steps: S21: Calculate the local residual copper ratio of each layer at the location point based on the stack-up layout, material distribution characteristics and copper foil ratio distribution in the PCB design file; S22: The equivalent residual copper rate at each location point is calculated using a thickness-weighted average model; S23: Extract the material property feature parameters of each dielectric layer to construct a material property feature vector. The material property feature parameters include the conductivity parameters of the dielectric layer, the fiberglass cloth weaving angle, the resin content percentage, and the dielectric layer thickness. S24: The equivalent residual copper rate and material property feature vector of the coordinate position point are fused to obtain a comprehensive feature vector; S25: The improved density peak clustering algorithm is used to perform cluster analysis on the comprehensive feature vector, dividing the PCB board into multiple anisotropic sub-back-drill regions.
5. The method for controlling the drilling depth of a back-drilled hole in a printed circuit board according to claim 4, characterized in that: In S22, the formula for calculating the equivalent residual copper rate is: ； In the formula, Location point The equivalent residual copper rate, Let be the theoretical design thickness of the i-th layer, and n be the total number of layers from the back-drilling start layer to the target signal layer. Let be the local residual copper ratio of the i-th layer at the location point.
6. The method for controlling the drilling depth of a back-drilled hole in a printed circuit board according to claim 4, characterized in that: S25 specifically includes the following sub-steps: S251: Calculate the weighted Euclidean distance between any two locations, and calculate the local density and relative distance of each location based on the weighted Euclidean distance; S252: Select the location points that simultaneously satisfy the local density greater than the local density threshold and the relative distance greater than the relative distance threshold as cluster centers. The two thresholds are adaptively determined by the inflection point method of the cluster decision graph. S233: Assign non-cluster center locations to the regions containing the nearest cluster centers to divide the initial sub-back-drill regions; S254: Morphological closure operation is used to perform boundary smoothing optimization on the initial sub-backdrill region to determine each sub-backdrill region; S255: Output the results of each sub-drilling region division, including the coordinate boundaries and region identifiers of each sub-region.
7. The method for controlling the drilling depth of back-drilling holes in a printed circuit board according to claim 1, characterized in that: In S251, the formula for calculating the weighted Euclidean distance is: ； In the formula, Let be the weighted Euclidean distance between point p and point q. , These are the normalized values of the k-th dimension of the comprehensive feature vector between positions p and q. denoted as the weight coefficient of the k-th dimension feature.
8. The method for controlling the drilling depth of back-drilling holes in a printed circuit board according to claim 1, characterized in that: In S3, the region thickness characteristic values include the region thickness standard deviation, region thickness skewness, and region thickness kurtosis.
9. The method for controlling the drilling depth of a back-drilled hole in a printed circuit board according to claim 7, characterized in that: In S4, the thickness compensation coefficient for each sub-back drill region is calculated using a weighted model, including the following steps: S41: Calculate the material property compensation factor for the sub-back drill region based on the eigenvalues in the normalized material property eigenvector. S42: Calculate the thickness feature compensation factor of the sub-back drill region based on the normalized thickness feature values of the three regions; S43: By using a preset two-dimensional weighted model, the material property compensation factor and the thickness feature compensation factor are fused together to calculate the thickness compensation coefficient of each sub-back drill region.
10. The method for controlling the drilling depth of a back-drilled hole in a printed circuit board according to claim 1, characterized in that: S5 specifically includes the following sub-steps: S51: For each sub-back-drilled area, based on the average thickness of the area and combined with the stacking theory parameters in the PCB design file, calculate the actual effective thickness from the starting layer to the target signal layer. S52: Based on the actual effective thickness of the sub-region, the thickness compensation coefficient, and the preset residual pile length, the back-drilling depth of each sub-back-drilling region is calculated; S53: Perform back-drilling depth safety threshold verification to determine the back-drilling depth of each sub-region.