High-precision micro blind hole processing method for optical module

Abstract

This application relates to the field of printed circuit board technology, and in particular to a high-precision micro-blind via processing method for optical modules. The method includes: acquiring the stack-up structure information of the substrate of the optical module to be processed, the position of the target conductive layer, and the micro-blind via design requirements; establishing a target via digital template corresponding to the via position to be processed; performing pre-forming rough processing based on the target via digital template while simultaneously acquiring online feature signals; identifying material layers and determining the remaining thickness state based on the online feature signals; switching to a low-damage interlayer transition processing state when entering the target termination region; comparing the actual via information with the target via digital template to generate a compensation control quantity; performing fine-tuning processing based on the compensation control quantity to complete the micro-blind via termination depth control and via shape correction, and performing post-processing corresponding to the via wall quality state. This application solves the problem that it is difficult to simultaneously achieve ideal conditions for via depth endpoint control and via shape consistency during micro-blind via processing.

Description

Technical Field

[0001] This application relates to the field of printed circuit board technology, and in particular to a method for processing high-precision micro blind holes for optical modules. Background Technology

[0002] As a key component in modern communication systems, the manufacturing process of the internal circuit board and packaging substrate of optical modules has a significant impact on communication performance. In high-density circuit boards, substrates, or substrate-like structures used in optical modules, micro-blind vias are crucial for achieving interlayer conductivity, and their processing quality directly affects the electrical performance and reliability of the device. Currently, micro-blind via processing mainly employs laser drilling technology. By setting specific energy parameters, the insulating layer is removed to form a hole structure for subsequent metallization, via filling, or interlayer conductivity. During processing, the system determines the target hole depth based on preset stack-up structure information and executes the processing according to fixed parameters. For high-density arrayed micro-blind vias, uniform processing parameters are typically used for batch processing. After processing, quality inspection is performed, and post-processing is conducted when necessary to improve the hole wall quality.

[0003] However, in the process of micro blind hole processing, it is difficult to achieve ideal results simultaneously in hole depth control and hole shape consistency, resulting in significant differences between different hole positions. Summary of the Invention

[0004] This application provides a method for high-precision micro blind hole fabrication for optical modules to solve the above-mentioned problems. The method includes:

[0005] S1. Obtain the stacked structure information, target conductive layer position and micro-blind hole design requirements of the optical module substrate to be processed, and based on the stacked structure information, target conductive layer position and micro-blind hole design requirements, establish a target hole digital template corresponding to the hole position to be processed. The target hole digital template is used to characterize the target hole depth, hole opening forming boundary and hole bottom forming requirements.

[0006] S2. Based on the target hole pattern digital template, perform pre-forming rough machining on the hole to be processed, and simultaneously collect online feature signals characterizing the material removal state during the pre-forming rough machining process;

[0007] S3. Identify the material layer where the current hole is being processed based on the online feature signal, and determine the remaining bottom thickness of the current hole relative to the target hole depth; when the current hole is identified as entering the target termination area, switch the processing state of the current hole from the pre-forming rough processing state to the interlayer transition low-damage processing state.

[0008] S4. Extract the actual hole shape information of the current hole position, and compare the actual hole shape information with the target hole shape digital template to obtain the hole shape deviation of the current hole position, and generate a compensation control quantity for correcting the current hole position processing based on the hole shape deviation.

[0009] S5. Perform fine finishing and shaping on the current hole position according to the compensation control amount to complete the micro blind hole termination depth control and hole shape correction, and perform post-processing on the hole wall corresponding to the hole wall quality state after the fine finishing and shaping is completed.

[0010] The above technical solution provides a single-hole forming benchmark by establishing a target hole shape digital template, uses online feature signals to achieve dynamic perception of processing status, triggers processing mode switching based on layer identification and remaining bottom thickness determination, generates compensation control quantities by comparing the actual hole shape with the target template, and finally completes closed-loop control through fine-tuning and differentiated post-processing. Thus, the synergistic effect of the various newly added technical features enables micro blind holes to achieve high-precision control requirements in four dimensions: termination depth, hole shape, hole bottom integrity, and hole wall quality. This effectively solves the problems of depth loss of control, hole shape instability, interface damage, and decreased interconnection reliability caused by the response differences of multilayer heterogeneous materials in the existing technology.

[0011] Optionally, the step of establishing a target hole pattern digital template corresponding to the hole position to be processed includes:

[0012] Read the stack-up structure file corresponding to the hole location, extract the insulation layer thickness, target conductive layer location, interlayer interface location, and wiring boundary information around the hole location, and create a template index according to the hole location;

[0013] The target hole pattern digital template includes at least three of the following: target hole depth, hole opening contour boundary, hole bottom retention area, and hole wall transition constraint, to form a hole pattern description set corresponding to a single hole location.

[0014] The above technical solution extracts the insulation layer thickness, target conductive layer position, interlayer interface position, and wiring boundary information around the hole by reading the stacked structure file. A unique template index is established according to the hole position. Then, at least three of the following are combined to form a hole type description set: target hole depth, hole opening contour boundary, hole bottom retention area, and hole wall transition constraint. This achieves a structured mapping from board-level structural data to single-hole-level forming specifications. With the help of this template, step S3 can determine the termination time based on the interlayer interface position and hole bottom retention area. Step S4 can perform deviation quantification based on the hole opening contour boundary and target hole depth. Step S5 can guide the generation of compensation path based on hole wall transition constraint. Thus, without introducing new hardware, the adaptability of micro blind hole processing to multi-layer heterogeneous structures and the closed-loop control accuracy are improved.

[0015] Optionally, the online feature signal includes at least two of the following: reflection response signal, plasma radiation signal, acoustic emission signal, local thermal response signal, and coaxial image signal;

[0016] During the pre-forming roughing process, the online feature signals are collected according to the processing pulse segment or scanning circle, and at least two of the following are extracted: signal amplitude, change slope, abrupt change position, thermal decay sequence and hole edge information, as the removal status input of the current hole position.

[0017] Through the above technical solution, at least two types of heterogeneous online feature signals are synchronously acquired during the pre-forming roughing process according to the processing pulse segment or scanning circle, and at least two of the amplitude, slope, abrupt change position, thermal decay sequence, and orifice edge information are extracted as removal state input quantities. This achieves multi-scale, multi-modal, and rhythmic synchronous perception of the material removal process. By leveraging the sensitivity of the reflection response signal to the surface penetration state, the responsiveness of the acoustic emission signal to the interface fracture behavior, the intuitiveness of the coaxial image signal for geometric forming, and the ability of the thermal response signal to characterize the energy deposition boundary, the signals form a complementary verification relationship at different processing stages. On this basis, through a structured acquisition strategy at the pulse segment or circle level, it is ensured that the extracted input quantities are strictly aligned with the actual processing actions, thereby providing a high-quality input foundation with temporal resolution, physical interpretability, and engineering feasibility for material layer identification as described above.

[0018] Optionally, identifying the material layer where the current hole is being processed based on the online feature signal includes:

[0019] Pre-establish characteristic response rules for the main insulating layer region, the interlayer interface transition region, and the bottom layer region near the target termination;

[0020] During the processing, the currently acquired online feature signals are matched with the feature response rules, and the material layer of the current hole position is determined by combining the previous hole processing records of the area to which the current hole position belongs.

[0021] Through the above technical solution, a reusable and calibrable criterion knowledge system is formed by pre-establishing characteristic response rules corresponding to three types of material layers. By using real-time matching of online feature signals and rules, semantic-level recognition of the material removal process is achieved. Combined with the layer recognition results of previous holes in the same processing sub-region, the regional processing history is introduced as a context correction factor to effectively suppress single-point signal drift caused by local heat accumulation, residue accumulation, or micro-deformation of the plate surface, thereby improving the robustness and engineering adaptability of material layer recognition. On this basis, a reliable premise is provided for the accurate determination of the remaining bottom thickness state and the switching of processing state in Example 5, ultimately supporting the high-precision forming of micro-blind holes in multi-layer heterogeneous structures with controllable endpoints, non-destructive interfaces, and consistent hole shapes.

[0022] Optionally, determining the remaining bottom thickness of the current hole position relative to the target hole depth includes:

[0023] At least two of the changes in reflection response, thermal response, and acoustic emission are combined to generate a termination proximity determination result.

[0024] When the termination proximity determination result enters the preset switching interval, the processing trajectory and processing mode corresponding to the pre-forming rough processing are stopped, and the processing trajectory, action path and pulse scheduling mode corresponding to the interlayer transition low damage processing state are called instead.

[0025] By combining at least two of the changes in reflection response, thermal response, and acoustic emission, a quantifiable termination proximity determination result is generated. When this result enters a preset switching interval, the processing trajectory, action path, and pulse scheduling mode are switched simultaneously, achieving active identification and flexible response to the endpoint region of micro-blind hole processing. Multi-physics signal fusion enhances the robustness of judging the proximity state of heterogeneous material layer interfaces, thus avoiding premature or delayed switching due to misjudgment of a single signal, ensuring the target hole depth termination accuracy and the integrity of the underlying structure. Furthermore, through the coordinated switching of trajectory, path, and scheduling mode, energy input is precisely matched to material removal requirements, suppressing defects such as hole bottom overheating, interface tearing, and recast layer thickening, providing a high-quality hole profile foundation for subsequent fine finishing and post-processing.

[0026] Optionally, extracting the actual hole type information of the current hole position includes:

[0027] Obtain at least two of the following information for the current hole location: hole profile, hole depth estimation, hole wall transition trend, and hole edge integrity.

[0028] The acquired information is then matched with the hole outline boundary, hole depth termination position, hole wall transition constraint, and hole edge constraint in the target hole type digital template to obtain the deviation of the current hole position in the hole opening, hole depth, hole wall, and edge items.

[0029] By comparing the orifice contour information with the orifice contour boundary, the orifice depth estimation information with the orifice depth termination position, the orifice wall transition trend information with the orifice wall transition constraint, and the orifice edge integrity information with the orifice edge constraint, a multi-dimensional deconstruction and structured error identification of the current orifice forming state is achieved. On this basis, each deviation item corresponds to a different correction action in the fine finishing process—the orifice expansion deviation item triggers the call of the circular trimming trajectory, the orifice wall transition too steep deviation item triggers the generation of a gradually shrinking spiral path, and the edge integrity missing deviation item triggers the edge strengthening and dwelling strategy. This ensures that the generation of compensation control quantities has clear physical meaning and is executable, avoids secondary damage caused by blind trimming, and improves the overall orifice shape consistency and processing controllability of the micro blind hole array.

[0030] Optionally, generating the compensation control quantity for correcting the current hole machining process based on the hole shape deviation includes:

[0031] According to the shaping rules corresponding to the deviation, determine the shaping direction, shaping sequence and shaping intensity of the current hole position, and generate the compensation control quantity corresponding to the current hole position;

[0032] The compensation control quantity acts on at least one of the following: processing energy distribution, scanning path composition, focus position calling, dwell order, and shaping repetition strategy, and is stored and called independently according to hole position.

[0033] Through the above technical solution, the deviation obtained by comparing the actual hole shape information with the target hole shape digital template is mapped into a shaping command with clear spatial orientation, execution timing and effect intensity. The compensation control quantity generated by the command is precisely bound to adjustable parameter dimensions such as processing energy, path, focus, dwell and repetition strategy, and then stored and recalled independently on a hole position basis, realizing a closed loop of the whole link from measurement deviation to process execution. On this basis, by means of spatial constraints of shaping direction, logical priority of shaping sequence and quantitative control of shaping intensity, shaping conflicts caused by multi-dimensional deviation coupling are effectively avoided, ensuring that each micro blind hole can obtain a customized shaping response that strictly matches its individual deviation characteristics without interfering with adjacent holes. Thus, in high-density array processing, the hole depth termination accuracy, hole mouth forming quality and hole wall transition consistency are stably maintained.

[0034] Optionally, after the compensation control quantity is used for the fine finishing and shaping of the current hole position, it is also written into the control record of the corresponding processing sub-area according to the area index;

[0035] Before processing subsequent holes, the layer determination result, termination proximity determination result, hole shape deviation result and compensation call result of the previous holes in the same processing sub-area are read. The compensation control quantity is inherited, modified or limited according to the positional relationship between holes before being used for the fine finishing and shaping of subsequent holes.

[0036] Through the above technical solution, by writing the compensation control quantity into the processing sub-area control record according to the regional index, and inheriting, correcting or limiting the layer determination result, termination proximity determination result, hole shape deviation result and compensation call result of the previous hole position according to the positional relationship between the holes, the structured accumulation and spatial transfer of regional process experience are realized. With the help of processing sub-area division and spatial proximity modeling, the subsequent holes can obtain an initial compensation benchmark with physical rationality before the complete online feature signal is collected. On this basis, combined with dynamic correction of thermal accumulation level, experience mistransfer is avoided. Finally, without increasing the perception burden of a single hole, the overall performance of the whole board micro blind hole array in terms of hole depth control accuracy, hole shape geometric consistency and thermally induced system deviation suppression is improved.

[0037] Optionally, during the low-damage processing state of interlayer transition and the finishing and shaping process, micro-environment linkage control is also implemented in the current hole processing area;

[0038] The microenvironment linkage control includes at least one of directional gas purging, local negative pressure extraction, micro-mist cooling, and liquid film sludge suppression.

[0039] Based on the current processing stage of the hole, the state of the hole profile, and the migration direction of the residue, switch at least one of the following: purging direction, extraction position, cooling intervention sequence, or liquid film coverage method.

[0040] Through the above technical solutions, by embedding microenvironment linkage control into the low-damage processing state of interlayer transition and the fine finishing process, and dynamically switching the purging direction, extraction position, cooling intervention sequence or liquid film coverage method according to the processing stage, orifice contour state and residue migration direction, it is possible to actively guide the migration path of molten material, accurately constrain local thermal boundaries, and enhance the physical stability of orifice edges. On this basis, by flexibly configuring at least one of the following technical means: directional gas purging, local negative pressure extraction, micro-mist cooling and liquid film slag suppression, the thickness of the recast layer inside the hole, the particle adhesion rate of the hole wall and the burr occurrence rate of the orifice are significantly reduced, and the geometric consistency and surface quality level of the micro-blind hole structure are improved, providing a process foundation for the reliable metallization and long-term service stability of high-density circuit boards for optical modules.

[0041] Optionally, the post-processing performed on the hole wall after the finishing and shaping process, corresponding to the hole wall quality state, includes:

[0042] First, the hole wall quality of the current hole location is inspected, and the corresponding post-processing path is selected from low-temperature plasma cleaning, mild chemical decontamination, microfluidic rinsing, ultrasonic-assisted cleaning and surface activation treatment according to the inspection results;

[0043] After post-processing, the hole wall quality results for that hole location are written into the subsequent compensation control record for similar holes or holes in the same area.

[0044] By using the above technical solutions as the basis for post-processing path decisions, a paradigm shift from fixed processes to on-demand responses is achieved. Furthermore, by writing the quality results into the subsequent compensation control record, post-processing is no longer isolated from the processing closed loop, but becomes a data source for optimizing front-end processing parameters. On this basis, combined with the indexing mechanism of similar holes and holes in the same area, the quality feedback is ensured to have process comparability and spatial transferability. Finally, through the mapping relationship between multi-dimensional quality indicators and multi-path post-processing, and the reverse adjustment effect of quality data on compensation control quantities, the problem of hole wall quality fluctuation caused by material heterogeneity, thermal accumulation differences, and poor local slag removal in optical module micro-blind holes is systematically solved, significantly improving the stability of subsequent metallization processes and the long-term reliability of devices. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art 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.

[0046] Figure 1 This is a flowchart illustrating a high-precision micro-blind hole fabrication method for an optical module, provided as an embodiment of this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, 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, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0048] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0049] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0050] Example 1

[0051] Based on the above issues, see Figure 1 As shown in the figure, a high-precision micro blind hole processing method for optical modules is provided in an embodiment of this application. The method includes the following steps:

[0052] Step 1: Obtain the stack-up structure information, target conductive layer position and micro-blind via design requirements of the optical module substrate to be processed, and based on the stack-up structure information, target conductive layer position and micro-blind via design requirements, establish a target hole digital template corresponding to the hole position to be processed. The target hole digital template is used to characterize the target hole depth, hole opening forming boundary and hole bottom forming requirements.

[0053] Among them, the target hole pattern digital template is a digital hole pattern benchmark model constructed for a single micro-blind hole position. Its technical connotation is: to record and encapsulate the three-dimensional geometric and forming quality constraints that the hole position is expected to achieve in the form of structured data.

[0054] The target hole depth can refer to the vertical distance from the surface to be processed to the upper surface of the target conductive layer. Its function is to provide a basis for determining the termination depth for subsequent processing.

[0055] The orifice forming boundary can refer to the contour envelope of the orifice on a horizontal cross section. Its function is to limit the orifice diameter, roundness, edge transition width and collapse tolerance range.

[0056] The requirements for hole bottom forming can refer to the diameter range of the hole bottom area, the surface flatness level, the allowable range of residual bottom thickness, and the interface integrity constraints. Their purpose is to ensure that the underlying metal conductor is not damaged and has the physical contact basis required for subsequent metallization.

[0057] This application uses the above three elements to form a target hole pattern digital template, so that each hole to be processed has an independent, comparable, and executable digital forming target.

[0058] Step 2: Perform pre-forming rough machining on the hole to be machined based on the target hole pattern digital template, and simultaneously collect online feature signals characterizing the material removal state during the pre-forming rough machining process;

[0059] Among them, pre-forming rough processing can refer to the use of laser processing with higher single-pulse energy, larger scanning radius and faster scanning rate to remove the main material of the target hole to quickly form a pre-blind hole structure; its function is to efficiently complete most of the removal of the target hole depth, leaving a controllable margin for subsequent layer identification and fine shaping;

[0060] Online feature signals can refer to physical response signals that are generated in real time during the processing and can reflect the current dynamic behavior of material removal; their role is to serve as the original input for processing status perception, supporting subsequent layer identification and bottom thickness determination.

[0061] This application, for example, determines whether material removal has entered the interface transition region based on the amplitude decay trend of the reflected response signal; this application, for example, determines whether the current processing is close to the metal layer based on the change in the spectral ratio of the plasma radiation signal; further, this application determines whether the current removal is in an unstable state based on the combination of the abrupt change frequency of the acoustic emission signal and the decay time sequence of the local thermal response signal. This application obtains a multidimensional characterization of the material removal state based on any of the above methods, thereby supporting the accurate determination of the subsequent processing state.

[0062] Step 3: Identify the material layer where the current hole is being processed based on the online feature signal, and determine the remaining bottom thickness of the current hole relative to the target hole depth; when the current hole is identified as entering the target termination area, switch the processing state of the current hole from the pre-forming rough processing state to the interlayer transition low-damage processing state.

[0063] Among them, the material layer can refer to the physical structure layer inside the substrate corresponding to the current laser beam action area, including the main insulating layer area, the interlayer interface transition area, and the bottom layer area near the target termination; its function is to provide spatial positioning basis for dynamic adjustment of processing parameters;

[0064] The remaining bottom thickness state can refer to the numerical range of the vertical distance between the current hole bottom and the upper surface of the target conductive layer, which serves as a decision condition for triggering the switching of the processing mode.

[0065] The target termination area can refer to the near termination judgment interval formed by extending a preset thickness threshold downwards from the target hole depth; its function is to provide early warning of the risk of the bottom layer approaching and to avoid over-processing.

[0066] This application, for example, identifies the current entry into the interlayer interface transition zone based on the step-like attenuation characteristics of the reflection response signal; this application, for example, identifies the current entry into the near-target termination bottom layer zone based on the combination of a sudden drop in the temperature rise rate of the thermal response signal and a sudden increase in the high-frequency energy of the acoustic emission signal; further, this application outputs a termination proximity determination result based on a weighted fusion of at least two of the changes in reflection response, thermal response, and acoustic emission. This application obtains the identification result of the current processing layer position and remaining bottom thickness state based on any of the above methods, and accordingly, when the target termination region is identified, it automatically stops calling the processing trajectory and processing mode corresponding to the pre-forming rough processing, and instead calls the processing trajectory, action path, and pulse scheduling mode corresponding to the low-damage interlayer transition processing state.

[0067] Step 4: Extract the actual hole shape information of the current hole position, and compare the actual hole shape information with the target hole shape digital template to obtain the hole shape deviation of the current hole position, and generate a compensation control quantity based on the hole shape deviation to correct the machining process of the current hole position.

[0068] The actual hole shape information can refer to the measured forming parameters of the current hole position in the spatial dimension obtained by non-contact measurement or image recognition after pre-forming rough processing or interlayer transition processing; its role is to provide an objective basis for comparison with the target template.

[0069] Hole shape deviation can refer to the numerical or morphological difference between the actual hole shape information and the corresponding item in the target hole shape digital template; its function is to quantify the degree to which the current hole shape deviates from the ideal state.

[0070] Compensation control quantity can refer to the set of correction instructions used to drive subsequent finishing and shaping processes, including energy, path, focus, dwell, or repetition strategies; its function is to achieve adaptive shaping of a single hole.

[0071] This application, for example, determines the scanning radius compensation value for annular trimming based on the deviation between the current aperture contour information and the target aperture contour boundary; this application, for example, determines the defocus compensation value for focal position retrieval based on the difference between the aperture depth estimation information and the target aperture depth; further, this application determines a combined compensation strategy for trimming sequence and trimming intensity based on the matching degree between aperture wall transition trend information and aperture wall transition constraints. This application obtains the compensation control amount for the current aperture position based on any of the above methods, and stores and retrieves them independently for each aperture position.

[0072] Step 5: Perform fine finishing and shaping on the current hole position according to the compensation control amount to complete the micro blind hole termination depth control and hole shape correction, and perform post-processing on the hole wall corresponding to the hole wall quality state after the fine finishing and shaping is completed;

[0073] Among them, fine finishing and shaping can refer to the fine hole shape control operation performed on the basis of low-damage interlayer transition processing, based on the compensation control amount; its function is to correct the hole depth endpoint and the geometric deviation of the hole opening / hole wall in a closed loop.

[0074] Post-processing corresponding to the pore wall quality status can refer to a differentiated treatment process that selects an appropriate path from various cleaning and activation methods based on the pore wall inspection results; its role is to improve the pore wall cleanliness and subsequent metallization bonding force.

[0075] This application, for example, calls a tapered spiral shaping trajectory based on the compensation control amount to continuously and smoothly correct the taper deviation from the orifice to the bottom of the orifice; this application, for example, calls a circular trimming trajectory at the orifice based on the compensation control amount to eliminate edge burrs and local edge enlargement; furthermore, this application inserts short pause segments based on the compensation control amount to achieve thermal balance trimming. This application completes the control of the micro-blind hole termination depth and the hole shape correction based on any of the above methods;

[0076] For example, in this application: after completing the low-damage interlayer transition machining, the system extracts the orifice diameter (38.2 μm), estimated orifice depth (92.5 μm), orifice edge defect amount (0.8 μm), and orifice wall cleanliness characterization value (medium) of the current orifice position. These values ​​are then compared item by item with the corresponding target orifice depth (95.0 μm), orifice contour boundary (±0.5 μm), orifice bottom retention area (≥2.0 μm), and orifice wall transition constraint (surface roughness Ra≤0.3 μm) in the target orifice digital template, generating a focal length shifted upwards by 1. The compensation control parameters were 2μm, 4.5μm reduction in scanning radius, 15% extension of dwell time, and two repeated shaping operations. Subsequently, fine finishing and shaping were performed to bring the final hole depth to 94.8μm, correct the hole diameter to 37.6μm, eliminate edge defects, and reduce the hole wall roughness to Ra=0.28μm. After processing, the hole wall was cleaned with low-temperature plasma to remove recast layer residue and microparticles. The hole wall quality score was improved to Grade A after cleaning. The hole wall quality result was recorded in the compensation control record of subsequent hole positions in the same processing sub-region.

[0077] Example 2

[0078] In an optional embodiment, this application also provides a digital template for creating a target hole pattern corresponding to the hole to be processed, including:

[0079] Step 1: Read the stack-up structure file corresponding to the hole location, extract the insulation layer thickness, target conductive layer location, interlayer interface location, and wiring boundary information around the hole location, and create a template index according to the hole location;

[0080] The stack-up structure file is a structured data file containing the material type, thickness, dielectric constant, conductor layer position, drill layer definition, and interlayer alignment reference for each layer of the substrate. Its function is to provide a physical configuration basis for the digital template, enabling the template construction process to be independent of manual experience input and possessing reproducibility and automation. The insulating layer thickness refers to the actual designed thickness of the resin insulating layer, PP film layer, or ABF film layer penetrated by the micro-blind via in the region where the via is located. This value directly affects the laser energy penetration depth and heat diffusion path, serving as the initial constraint condition for the termination calculation of the via depth in the template. The target conductive layer position refers to the theoretical position of the metal conductor layer (e.g., copper foil layer or redistribution layer) that the micro-blind via needs to ultimately connect in the Z-axis coordinate system. This position, together with the insulating layer thickness, determines the geometric reference for the target via depth, used to calibrate the processing endpoint. The interlayer interface position refers to the Z-axis coordinate at the junction of two adjacent material layers, including the insulating layer / The interfaces of metal layers, different resin layers, or reinforcing materials / matrix serve to identify areas of material abrupt change that require key responses during subsequent processing, providing spatial positioning for layer identification and transition processing mode switching. The wiring boundary information around the hole refers to the spatial distribution range of signal lines, power planes, heat dissipation pads, or shielding structures within the vicinity of the hole. Its function is to constrain the forming boundary of the hole opening and the transition shape of the hole wall, preventing deformation, short circuits, or impedance abnormalities in adjacent lines due to the expansion of the heat-affected zone. Establishing a template index by hole position involves assigning a unique identifier (e.g., a string ID generated based on XY coordinate hashing) to each hole to be processed, and binding all the previously extracted parameters to this ID to form an independent, callable data unit. Its function is to support differentiated modeling and closed-loop control at the single-hole level, allowing holes in different areas and with different functional requirements on the same substrate to call their respective suitable templates, avoiding forming deviations caused by parameter uniformity.

[0081] This application, for example, reads IPC-2581 format files using a layer stacking structure file parsing engine, automatically identifies and extracts the thickness fields of each layer and the layer stacking order, and locates the Z coordinate of the target conductive layer using a design layer number mapping table; this application, for example, obtains the inter-layer interface position by parsing the LayerStackup section of the ODB++ file, and uses Gerber graphic layer overlay analysis to obtain the wiring boundary around the via; furthermore, this application uses a coordinate mapping algorithm to spatially align the physical via coordinates with the layer stacking structure data, generating a template index entry with a unique ID. This application obtains the foundation for constructing a target via digital template that precisely matches a single via based on any of the above methods.

[0082] For example, this application involves: for a micro-blind via in a high-speed interconnect area of ​​an optical module carrier board, the system reads its corresponding ODB++ stack-up structure file, identifies that the via is located between the 3rd and 4th layers, needs to penetrate a 50μm thick ABF insulating film, the target conductive layer is the 4th layer copper foil, and its Z-axis position is 82μm from the surface; the interlayer interface positions include the ABF / copper interface (Z=82μm) and the upper FR-4 / ABF interface (Z=32μm); a differential signal line with a width of 8μm is adjacent to the right side of the via, only 45μm from the center of the via; accordingly, the system generates a template index IDH03-08245 for the via and writes the above five parameters into the data structure associated with the ID as the reference input for all subsequent processing control.

[0083] Step 2: The target hole pattern digital template includes at least three of the following: target hole depth, hole opening contour boundary, hole bottom retention area, and hole wall transition constraint, to form a hole pattern description set corresponding to a single hole location;

[0084] The target hole depth refers to the vertical distance from the substrate surface to the upper surface of the target conductive layer. Its value is determined by the insulating layer thickness and the target conductive layer position, and is used to set the main control dimension for processing termination. In this embodiment, this parameter directly participates in the calculation of the threshold for determining the remaining bottom thickness state in step S3. The hole contour boundary refers to the maximum allowable circumscribed circle diameter, the minimum inscribed circle diameter, and the allowable contour deviation bandwidth at the micro-blind hole opening. Its function is to limit the upper limit of the hole forming quality and prevent excessive hole diameter expansion due to thermal collapse or molten splashing, which would affect subsequent pattern alignment and... The accuracy of hole filling; the hole bottom retention area can refer to the range of insulation layer thickness allowed to be retained above the target conductive layer, such as 2μm to 5μm. Its function is to avoid overheating of the underlying metal, interface tearing and damage to the conductive layer, and to provide a basis for identifying the target termination area in step S3; the hole wall transition constraint can refer to the taper change rate, curvature continuity requirement or local transition zone length limit of the hole wall from the hole opening to the hole bottom. Its function is to guide the adjustment direction of the compensation control amount on the scanning path and focus position in steps S4 / S5, and to ensure that the interface area is removed smoothly and the recast layer is controllable;

[0085] Any combination of at least three of the above four technical features constitutes a complete set of hole shape descriptions; this set is not limited to geometric dimensions, but rather integrates material response characteristics, process constraints and packaging reliability requirements into a multi-dimensional forming specification; its function is to transform abstract design intent into an executable, comparable and compensable digital control object, supporting the item-by-item deviation identification between actual hole shape information and template in step S4.

[0086] This application provides, for example, a template for constructing three elements: target hole depth, hole opening contour boundary, and hole bottom retention area. This template is suitable for micro-blind holes in high-density signal areas where stringent requirements exist for hole depth consistency and bottom thickness safety margin. Another application provides a template for constructing three elements: target hole depth, hole opening contour boundary, and hole wall transition constraint. This template is suitable for micro-blind holes in areas near the power plane where higher requirements exist for hole opening edge integrity and hole wall smoothness. Furthermore, this application provides a template for all four elements, suitable for composite micro-blind holes in the optical module driver chip connection area where high-precision hole depth control, low edge collapse risk, strong interface protection, and high fill adaptability are simultaneously required. Based on any of the above combinations, this application obtains a target hole type digital template that is engineering-operable and meets specific hole position functional requirements.

[0087] For example, in this application: for the aforementioned hole with ID H03-08245, the system calls a four-item complete template, sets the target hole depth to 82μm±1μm, and the hole opening contour boundary is... The hole diameter is 85μm±3μm and the roundness is ≤0.8μm. The hole bottom retention range is 3μm to 4μm. The hole wall transition constraint is a taper of 6°±0.5° and the transition zone length is ≥15μm. This template is written into the hole position parameter library of the control unit and is loaded before the pre-forming rough machining of step S2 is started, serving as the reference source for hole shape deviation comparison in step S4.

[0088] Example 3

[0089] In one embodiment, this application also provides online feature signals including at least two of the following: reflection response signal, plasma radiation signal, acoustic emission signal, local thermal response signal, and coaxial image signal. During the pre-forming rough machining process, the online feature signals are collected according to the processing pulse segment or scanning circle, and at least two of the following are extracted: signal amplitude, change slope, abrupt change position, thermal decay sequence, and hole edge information, as the removal status input of the current hole position.

[0090] Step 1: Online feature signals include at least two of the following: reflection response signal, plasma radiation signal, acoustic emission signal, local thermal response signal, and coaxial image signal;

[0091] Among them, the reflection response signal (the signal characterized by the intensity of reflected light and its dynamic changes after laser incident on the surface), the plasma radiation signal (the signal characterized by the intensity of visible / near-infrared light and the intensity of spectral lines emitted by the plasma plume during the processing), the acoustic emission signal (the signal characterized by the high-frequency elastic waves excited by thermal stress and phase change impact on the material), the local thermal response signal (the signal characterized by the rate of temperature change, peak temperature rise and thermal decay time constant of the surface or near-surface layer of the hole area), and the coaxial image signal (the signal characterized by the real-time grayscale image of the hole opening, edge contour and brightness distribution acquired by the imaging system arranged coaxially with the processing optical path) are all typical physical response signals used in this field to characterize the dynamic behavior of material removal during laser micromachining.

[0092] In this embodiment, the reflection response signal is used to reflect the optical penetration state and surface morphology changes of the current processing interface; the plasma radiation signal is used to reflect the degree of material vaporization and the intensity of the plasma shielding effect; the acoustic emission signal is used to reflect the mechanical response behavior such as material fracture, spalling and interface peeling; the local thermal response signal is used to reflect the energy deposition efficiency and thermal diffusion boundary evolution; and the coaxial image signal is used to directly provide visual feedback on the orifice geometry. The above five types of signals have different response sensitivities and characteristic saliencies in different material layers and different processing stages. Their combined use can cover the characterization needs of the removal process from multiple dimensions, including optics, thermal, mechanics and image.

[0093] This application, for example, can identify the transition from the main insulating layer region to the interlayer interface transition region based on the coordinated change trend of the reflection response signal and the plasma radiation signal; it can also locate the critical moment approaching the target termination layer based on the matching relationship between the abrupt change position of the acoustic emission signal and the thermal decay time sequence of the local thermal response signal; further, it can determine the initial state of the orifice collapse based on the synchronous jump between the inflection point of the sharpness decrease of the orifice edge information in the coaxial image signal and the attenuation rate of the reflection response signal amplitude. Based on any of the above methods, this application obtains multidimensional inputs characterizing the current material removal state at the orifice location, providing a highly robust data foundation for subsequent layer identification and remaining thickness determination.

[0094] Step 2: During the pre-forming rough machining process, online feature signals are collected according to the machining pulse segment or scanning circle, and at least two of the following are extracted: signal amplitude, change slope, abrupt change position, thermal decay sequence and hole edge information, as the removal status input of the current hole position;

[0095] Among them, the processing pulse segment can refer to a time segment divided into units of a fixed number of laser pulses, such as 10–50 pulses constituting a pulse segment; the scanning circle can refer to the processing cycle in units of a single closed trajectory scan, such as one scan circle being completed in a spiral path; this acquisition strategy ensures that signal sampling and processing actions are strictly synchronized on the time scale, avoiding feature mismatch caused by asynchronous sampling.

[0096] Signal amplitude can refer to the average reflected light intensity of the reflected response signal in a certain pulse segment, the peak radiation intensity of the plasma radiation signal in a certain cycle, the root mean square value of the effective sound pressure of the acoustic emission signal in a certain pulse segment, the surface temperature rise value of the local thermal response signal at the end of a certain cycle, and the average gray value of the orifice region of the coaxial image signal in a certain pulse segment.

[0097] The slope of change can refer to the rate of change of the amplitude of the reflected response signal with the pulse segment number, the rise / fall slope of the intensity of the plasma radiation signal with the number of scans, the time derivative of the energy accumulation rate of the acoustic emission signal, the initial slope of the temperature rise curve of the local thermal response signal, and the linear fitting slope of the estimated value of the orifice diameter of the coaxial image signal with the number of scans.

[0098] The abrupt change position can refer to the pulse segment number where the amplitude of the reflection response signal drops sharply, the scanning cycle in which the spectral enhancement inflection point of the plasma radiation signal appears, the pulse segment in which the energy of the high-frequency component of the acoustic emission signal first exceeds the threshold, the cycle in which the thermal decay time of the local thermal response signal is significantly delayed, and the pulse segment in which the aperture edge blur index of the coaxial image signal first breaks through the set threshold.

[0099] Thermal decay timing can refer to the time required for the local thermal response signal to fall back from the peak temperature rise to the full width at half maximum (FWHM), or the relaxation time constant from the peak to the stable baseline.

[0100] The orifice edge information can refer to the set of pixel coordinates of the orifice contour, the distribution of gradient magnitude of the orifice edge, or the orifice edge integrity score extracted by the edge detection algorithm from the coaxial image signal;

[0101] In this embodiment, the signal amplitude reflects the energy coupling efficiency of the current removal stage; the slope of change reflects the acceleration or decay trend of the removal rate; the abrupt change location identifies key nodes where material properties or interface states change significantly; the thermal decay sequence reflects the balance between local heat accumulation and dissipation; and the orifice edge information directly characterizes the evolution of orifice geometry quality. Any combination of the above two items constitutes the removal state input, which can take into account both steady-state response and transient characteristics, supporting the reliable discrimination of the subsequent layer identification model.

[0102] This application can, for example, determine whether the current hole position is in the stable removal stage of the resin insulation layer based on the joint change of the amplitude and slope of the reflection response signal; it can also identify the interlayer interface crossing time based on the coupling relationship between the abrupt change position of the acoustic emission signal and the thermal decay sequence of the local thermal response signal; further, it can confirm the hole edge collapse risk level based on the consistency between the continuous degradation trend of the hole edge information in the coaxial image signal and the attenuation rate of the reflection response signal amplitude. This application obtains the removal status input of the current hole position based on any of the above methods, ensuring that it strictly corresponds to the processing action cycle and has the feature discrimination capability for material layer identification tasks.

[0103] For example, in the process of machining micro-blind holes on an AB-type multilayer substrate for an optical module, during the pre-forming rough machining stage, the system simultaneously collects reflection response signals and acoustic emission signals in pulse segments of 25 laser pulses each. In the fourth pulse segment, the amplitude of the reflection response signal decreases by 18% compared to the previous segment, and the slope of change changes from +0.3% / segment to −2.1% / segment. At the same time, the acoustic emission signal shows the first high-frequency energy surge peak in this segment, which corresponds to the 22nd pulse. Combining the prior information that the thickness of the first insulating layer in the substrate stacked structure is 65μm and the interface transition zone starts at 62μm from the surface, the control system determines that the current hole position has entered the interlayer interface transition zone and sends the three parameters corresponding to the pulse segment—amplitude, slope, and abrupt change position—as removal status inputs to the layer position identification module.

[0104] For example, this application can also be as follows: When performing micro blind hole processing on a rigid-flex optical module board, the system acquires coaxial image signals and local thermal response signals in single-turn spiral scans; at the end of the 7th scan, the coaxial image signal shows that the edge integrity score of the hole has dropped to 0.62 (out of 1.0), and the standard deviation of the edge gradient has increased by 47%; at the same time, the thermal decay time of the local thermal response signal has been extended from 1.8s in the previous turn to 2.9s; based on this, the system determines that the thermal accumulation in the hole area has intensified and is accompanied by edge deterioration, and uses the hole edge information and thermal decay time as the removal state input quantity to trigger the preparatory judgment of entering the low-damage processing state of interlayer transition in advance.

[0105] Example 4

[0106] In another optional embodiment, this application also provides for identifying the material layer where the current hole is being processed based on online feature signals, including:

[0107] Step 1: Pre-establish characteristic response rules for the main insulating layer region, the interlayer interface transition region, and the bottom layer region near the target termination layer, respectively;

[0108] Among them, the main insulating layer region can refer to the continuous thickness region in the optical module substrate that is far away from the interfaces of various materials and is mainly composed of homogeneous resin or glass cloth reinforcement medium; the interlayer interface transition region can refer to the region of nonlinear physical response change within ±5μm at the junction of two adjacent functional layers (such as the insulating layer and the metal conductor layer, or between different types of insulating layers); the near target termination bottom layer region can refer to the region of residual material to be removed that is ≤20μm away from the upper surface of the target conductive layer.

[0109] The characteristic response rules corresponding to the main area of ​​the insulating layer are: the amplitude of the reflection response signal is stable and the attenuation slope is gentle; the intensity of the plasma radiation signal is moderate and the spectral distribution is concentrated; the pulse interval of the acoustic emission signal is uniform and the proportion of high frequency components is low; the local thermal response signal has a constant temperature rise rate and a long thermal attenuation time; and the edge of the aperture in the coaxial image signal has high clarity and continuous contour.

[0110] The characteristic response rules corresponding to the interlayer interface transition zone are as follows: the reflection response signal shows a small abrupt change or slope reversal in stages; the intensity of the plasma radiation signal increases sharply and the proportion of short-wave energy increases; the frequency of acoustic emission signal pulses increases and the energy of high-frequency components jumps; the local thermal response signal temperature rise rate accelerates and the thermal decay time shortens; and the aperture edge in the coaxial image signal shows local blurring or uneven brightness.

[0111] The characteristic response rules corresponding to the near-target termination bottom layer region are as follows: the amplitude of the reflection response signal decreases significantly and is accompanied by high-frequency jitter; the intensity of the plasma radiation signal fluctuates violently and the spectral ratio shifts; the acoustic emission signal shows dense short-time pulse groups and the peak value of the impact response increases; the local thermal response signal shows a sharp increase in the temperature rise rate and a sharp shortening of the thermal decay time sequence; and the integrity of the aperture edge in the coaxial image signal decreases and the contour continuity is interrupted.

[0112] This application identifies the main insulating layer region based on a combined criterion of the stability of the reflection response signal amplitude and the trend of attenuation slope; it identifies the interlayer interface transition region based on a combined criterion of the increase in plasma radiation signal intensity and the change in the proportion of short-wavelength energy; and it identifies the bottom layer region approaching the target based on a combined criterion of the sudden increase in acoustic emission signal pulse density and the shortening ratio of the thermal attenuation time of the local thermal response signal. Based on any of the above methods, this application obtains the ability to characterize the differentiated responses of the three material layers, providing a structured criterion basis for subsequent signal matching.

[0113] Step 2: During the processing, the currently acquired online feature signals are matched with the feature response rules, and combined with the previous hole processing records of the area to which the current hole position belongs, the material layer position of the current hole position is determined.

[0114] Among them, the currently acquired online feature signals can refer to at least two of the following: reflection response signals, plasma radiation signals, acoustic emission signals, local thermal response signals, and coaxial image signals, which are synchronously acquired according to the processing pulse segment or scanning circle during the pre-forming roughing process; the previous hole processing record of the area to which the current hole position belongs can refer to at least one of the following: the layer identification result, the termination proximity determination result, the actual hole depth estimate, and the compensation control quantity call status of adjacent holes that have been processed in the same processing sub-area.

[0115] This application matches the characteristic response rules corresponding to the interlayer interface transition zone based on the decreasing trend of the amplitude of the currently acquired reflection response signal and the shortening ratio of the thermal decay time sequence. Combined with the fact that the region has already shown two interface recognition results in the previous hole processing record, this application confirms that the current hole position is in the interlayer interface transition zone. Based on the sudden increase in the pulse density of the currently acquired acoustic emission signal and the shift in the ratio of the plasma radiation signal spectrum, this application matches the characteristic response rules corresponding to the near-target termination bottom layer region. Combined with the fact that the average remaining bottom thickness in this region has been reduced to 12 μm in the previous hole processing record, this application determines that the current hole position has entered the near-target termination bottom layer region. Based on the increasing ambiguity of the hole edge and the reversal of the slope of the reflection response signal in the currently acquired coaxial image signal, this application matches the intermediate response mode of the evolution from the main insulating layer region to the interlayer interface transition zone. Referring to the fact that no interface recognition event has been triggered in this region in the previous hole processing record, this application determines that the current hole position is still at the end of the main insulating layer region. This application obtains a dynamic, context-aware determination result of the material layer position of the current hole position based on any of the above methods.

[0116] For example, this application takes a multilayer carrier board for an optical module as an example. Its stacked structure includes three FR-4 insulating layers and two copper conductor layers, with the target conductive layer being the second copper conductor layer. When performing micro-blind via processing on this board, the system first reads the stacked structure file and determines that the thicknesses of each insulating layer are 80μm, 65μm, and 70μm, respectively, and that the upper surface of the second copper conductor layer is 215μm from the starting processing surface. Based on this, the main insulating layer area (0–75μm, 75–140μm, 140–210μm), the interlayer interface transition area (75±5μm, 140±5μm, 215±5μm), and the near-target conductive layer are constructed. The system uses a three-category response rule library for the termination layer region (210–215 μm). When machining hole number 17, the system detects that the amplitude of the reflected signal decreases by 18% and the thermal decay time shortens by 37% after the 42nd scan. At the same time, holes number 12 and 15, which were previously machined in the sub-region to which this hole belongs, trigger interlayer interface identification in the same scan cycle. Based on this, the system matches the interlayer interface transition zone rule and determines that hole number 17 is currently in the 140±5 μm interface transition zone. This determination result directly triggers the focused evaluation of the remaining bottom thickness in subsequent steps and serves as a prerequisite for whether to initiate the low-damage interlayer transition machining state.

[0117] Example 5

[0118] In one possible implementation, this application also provides a method for determining the remaining bottom thickness state of the current hole position relative to the target hole depth, including:

[0119] Step 1: Combine at least two of the changes in reflection response, thermal response, and acoustic emission to generate a termination proximity determination result;

[0120] Among them, the change in reflection response can refer to the amplitude attenuation, slope abrupt change, or periodic oscillation characteristics of the light intensity signal reflected from the surface or inside the hole to the optical detection unit during laser processing as the processing depth changes; the change in thermal response can refer to the changes in the temperature rise rate, peak temperature, thermal diffusion delay, or thermal decay time constant of the local temperature field in the processing area under pulse action; the change in acoustic emission can refer to the amplitude, spectral energy distribution, pulse count density, or temporal abrupt change characteristics of the high-frequency mechanical vibration signal excited when the material undergoes microcracks, phase transitions, or interface peeling under the action of laser thermo-mechanical coupling.

[0121] The three changes mentioned above technically characterize the optical interface response, thermal conduction behavior, and mechanical fracture state during the material removal process, respectively. Their inherent function is to reflect the interaction strength and instability tendency between the current processing front and different material layers (such as resin insulation layer, glass fiber reinforcement layer, and copper metal layer). In this embodiment, the reflection response change is used to identify the enhanced interface reflection phenomenon when the bottom of the hole is close to the high reflectivity metal layer, the thermal response change is used to identify the phenomenon of increased local heat accumulation or decreased heat dissipation efficiency caused by abrupt changes in thermal conductivity, and the acoustic emission change is used to identify microscale energy release events before brittle fracture or interface tearing. Together, these three constitute the multidimensional perception basis for the state of the remaining bottom thickness approaching the critical threshold.

[0122] This application may determine the termination proximity assessment result based on, for example, a weighted fusion result of reflection response changes and thermal response changes; alternatively, it may determine the termination proximity assessment result based on a logical AND gate criterion of reflection response changes and acoustic emission changes; furthermore, it may determine the termination proximity assessment result based on a sliding window correlation analysis result of thermal response changes and acoustic emission changes. This application obtains a quantitative assessment basis for whether the current aperture position has entered the target termination region based on any of the above methods.

[0123] For example, in this application, when the pre-forming roughing process reaches the 82nd scan, the amplitude of the reflection response signal increases for three consecutive pulse cycles (increase ≥ 18%), while the temperature rise rate of the local thermal response signal decreases to 42% of the initial value, and the energy integral value of the acoustic emission signal in the 20–50kHz frequency band suddenly increases by 2.3 times; the control unit sums the three changes according to preset weighting coefficients (0.4, 0.35, 0.25) and outputs a termination proximity judgment result of 0.87; this value falls within the preset switching range [0.75, 1.0], triggering subsequent switching actions.

[0124] Step 2: When the termination proximity determination result enters the preset switching range, stop calling the processing trajectory and processing method corresponding to the pre-forming rough machining, and instead call the processing trajectory, action path and pulse scheduling method corresponding to the interlayer transition low damage machining state.

[0125] Among them, the preset switching interval is a range of technical parameters used to define the critical range where the degree of termination approaches and the protection mechanism needs to be activated. Its boundary value is determined based on historical processing data statistics and material layer response calibration experiments, and is not dynamically adjusted with each single processing. The low-damage interlayer transition processing state is an independent process state that is different from pre-forming rough processing. Its core is to suppress the thermal stress concentration and mechanical erosion effect when crossing the material layer interface. The processing trajectory can refer to the spatial movement path of the laser beam on the cross section of the hole area. The action path can refer to the spatial action distribution of laser energy in the hole depth direction. The pulse scheduling method can refer to the combination strategy of single pulse energy, pulse repetition frequency, pulse interval and pulse sequence arrangement.

[0126] In the above technical features, the preset switching interval has been defined and explained for the first time in Embodiment 4, and is only referred to here naturally; the processing trajectory and processing method corresponding to the pre-forming roughing have appeared as existing technical features in Embodiments 2, 3, and 4, and their meaning and implementation have been explained in the aforementioned embodiments, and will not be repeated in this step; the interlayer transition low-damage processing state has been introduced as a new technical feature in Embodiments 3 and 4, and its correspondence with the material layer identification result has been clarified in Embodiment 4, and this step only follows this definition without further elaboration; the processing trajectory, action path, and pulse scheduling method together constitute the execution carrier of the interlayer transition low-damage processing state, and its role in this step is: after the termination proximity determination result is triggered, it acts as the called object, replacing the original roughing parameter set, thereby enhancing the spatial focusing of processing energy, refining the time allocation, and weakening the thermo-mechanical coupling effect.

[0127] This application, for example, can use the condition of entering a preset switching interval based on the termination proximity determination result, call a pre-stored annular trimming trajectory as the processing trajectory, call the action path with a small defocus offset, and adopt a pulse scheduling method that reduces single pulse energy and increases pulse interval; this application can also use the same condition, call a tapered spiral trajectory as the processing trajectory, call the action path of center-edge segmented focusing, and adopt a pulse scheduling method with decreasing pulse energy gradient and dynamic frequency reduction; furthermore, this application can also use the same condition, call a low coverage contour scanning trajectory as the processing trajectory, call the action path of orifice wall transition zone directional focusing, and adopt a pulse scheduling method that inserts short pause segments. This application obtains controllable, low-disturbance, and non-destructive trimming capability for the bottom area of ​​the aperture based on any of the above methods.

[0128] For example, after the above-mentioned output termination proximity determination result is 0.87, the control system immediately interrupts the currently executing multi-turn extended roughing trajectory, reads the interlayer transition low-damage machining configuration file numbered LTD-05 from the parameter library, and loads the defined annular trimming trajectory (scanning radius of 92% of the orifice diameter, a total of 3 turns), action path (focus position moved 12μm higher than roughing, forming a +12μm defocusing amount), and pulse scheduling mode (single pulse energy reduced to 65% of roughing, pulse interval extended to 800ns); then, the subsequent machining is performed with this configuration until the interlayer transition stage is completed.

[0129] Example 6

[0130] In one possible implementation, this application also provides for extracting the actual hole type information of the current hole position, including:

[0131] Step 1: Obtain at least two of the following information for the current hole location: hole opening profile, hole depth estimation, hole wall transition trend, and hole opening edge integrity.

[0132] The orifice contour information can refer to an image or set of geometric parameters characterizing the shape, size, and edge continuity of the current orifice projected onto the XY plane. For example, it could be the orifice edge pixel coordinate sequence, fitted circle diameter, roundness error value, or edge grayscale gradient distribution acquired by a coaxial vision system. In this embodiment, it reflects the forming quality state of the orifice region after pre-forming rough machining and serves as a direct input for comparison with the orifice contour boundary in the target hole digital template. The hole depth estimation information can refer to an approximate removal depth along the Z direction of the current hole position obtained based on online feature signal inversion or geometric modeling. For example, it could be an estimated hole depth obtained based on the attenuation time of the reflected signal, the peak position of the thermal response, or the acoustic emission abrupt change depth mapping. In this embodiment, it characterizes the positional relationship between the current processing progress and the target hole depth and serves as the basis for determining the deviation from the hole depth termination position in the target hole digital template. The hole wall... The transition trend information can refer to data features describing the shape change law of the hole wall from the hole opening to the hole bottom, such as the curvature change rate, taper slope gradient, or local unevenness spectrum energy distribution of the hole wall contour point cloud along the depth direction. In this embodiment, it is used to identify whether the hole wall has local collapse, recast layer accumulation, or interface instability tendency, and to support the matching analysis with the hole wall transition constraint in the target hole type digital template. The hole opening edge integrity information can refer to qualitative or semi-quantitative indicators reflecting whether the hole opening edge has defects, burrs, melt adhesion, or local expansion, etc., such as the proportion of hole opening edge pixel breakage length in the coaxial image, the magnitude of edge sharpness reduction, or the proportion of edge disturbance energy in the high-frequency vibration signal. In this embodiment, it is used to evaluate the hole opening structure integrity, which is a key input to determine whether to perform ring trimming or edge strengthening processing, and corresponds to the hole opening edge constraint in the target hole type digital template.

[0133] This application may, for example, obtain orifice contour information using coaxial visual image recognition and geometric fitting methods; it may also obtain orifice depth estimation information using a joint inversion method based on the attenuation time sequence of the reflected signal and the peak position of the thermal response; further, it may obtain orifice wall transition trend information and orifice edge integrity information using orifice wall point cloud curvature gradient calculation and edge pixel breakage analysis methods. Based on any of the above methods, this application obtains multidimensional actual orifice shape parameters to support subsequent orifice shape deviation identification and compensation control quantity generation.

[0134] For example, after completing the low-damage interlayer transition processing, the control system calls the coaxial vision module to perform a focal plane scan on the current hole position to obtain a high-contrast image of the hole opening area; performs edge detection and circular / elliptical fitting on the image, and outputs the hole opening diameter, roundness error, and edge continuity score; simultaneously, the time point when the reflected signal recorded at the hole position in the pre-forming rough processing stage first shows significant attenuation is converted into the current hole depth estimate by combining the known laser pulse propagation speed and material group velocity correction coefficient; furthermore, the control system constructs a hole wall morphology trend model based on the frequency of abrupt changes in acoustic emission signals and the local thermal response attenuation delay collected during the previous processing, and outputs the hole wall taper deviation trend and edge collapse risk level; finally, any two or more of the above four pieces of information are combined as the actual hole shape information of the current hole position and input to the comparison module.

[0135] Step 2: Match the obtained information with the hole outline boundary, hole depth termination position, hole wall transition constraint and hole edge constraint in the target hole digital template to obtain the deviation of the current hole position in the hole opening, hole depth, hole wall and edge items;

[0136] The orifice contour boundary is the ideal geometric boundary of the orifice defined in the target orifice digital template, such as a set of constraints including the set orifice diameter range (120±5μm), allowable roundness tolerance (≤0.03), and edge transition width (0.5–1.2μm). In this embodiment, it serves as a comparison benchmark for orifice contour information to identify whether the current orifice is out of tolerance, too large, too small, or out of round. The orifice depth termination position is the Z-coordinate of the target orifice bottom or the allowable residual bottom thickness range (5±1μm) set in the target orifice digital template. In this embodiment, it serves as a comparison benchmark for orifice depth estimation information to identify whether the current orifice depth has not reached, just reached, or exceeded the target termination depth. The orifice wall transition constraint is... The technical requirements set in the target hole pattern digital template for the morphological changes of the hole wall from the hole opening to the hole bottom, such as allowable taper range (4°–8°), maximum rate of curvature change (≤0.15μm / μm²), and no local depressions, are used as a comparison benchmark for hole wall transition trend information in this embodiment to identify whether there are transition discontinuities, local collapses, or recast layer accumulation. The hole opening edge constraint is a limitation on the structural integrity of the hole opening edge in the target hole pattern digital template, such as edge sharpness threshold (≥0.8), upper limit of defect length (≤3μm), and no molten adhesion, etc. In this embodiment, it is used as a comparison benchmark for hole opening edge integrity information to identify whether edge trimming action needs to be initiated.

[0137] This application can, for example, compare the fitted circle diameter in the orifice contour information with the set orifice diameter range in the orifice contour boundary, and mark the orifice expansion deviation when the measured diameter exceeds the upper limit; this application can also, for example, perform interval inclusion judgment between the estimated value in the orifice depth estimation information and the allowable residual bottom thickness interval in the orifice depth termination position, and mark the orifice depth insufficient deviation when the estimated value is less than the lower limit; further, this application can also compare the peak curvature gradient in the orifice wall transition trend information with the maximum curvature change rate in the orifice wall transition constraint by a threshold, and mark the orifice wall local collapse deviation when it exceeds the threshold. Based on any of the above methods, this application obtains a structured list of deviations for the current orifice position in four dimensions: orifice, orifice depth, orifice wall and edge, with each item clearly indicating the specific deviation type, location, and severity level.

[0138] For example, this application may compare the obtained orifice diameter of 126.3 μm with the orifice contour boundary (120 ± 5 μm) set in the target orifice digital template, and determine it as an orifice expansion deviation item with a deviation of +1.3 μm; compare the estimated orifice depth of 4.2 μm with the orifice depth termination position (5 ± 1 μm), which falls within the allowable range, and no deviation item is generated; compare the peak value of the orifice wall taper slope gradient of 0.18 μm / μm² with the orifice wall transition constraint (≤0.15 μm / μm²), and determine it as an orifice wall transition too steep deviation item; compare the orifice edge fracture length ratio of 12% with the orifice edge constraint (≤5%), and determine it as an edge integrity missing deviation item. Finally, a structured list containing the three deviation items is formed for subsequent compensation control quantity generation module to call.

[0139] Example 7

[0140] In one embodiment, this application also provides a method for generating a compensation control quantity based on hole shape deviation, including:

[0141] Step 1: According to the shaping rules corresponding to the deviation, determine the shaping direction, shaping sequence and shaping intensity of the current hole position, and generate the compensation control quantity corresponding to the current hole position;

[0142] Among them, deviation items can refer to the specific types of deviations identified in the four items of hole opening, hole depth, hole wall and edge after comparing the actual hole pattern information with the target hole pattern digital template.

[0143] The shaping rules are a mapping table pre-existing in the control unit. This mapping table associates each type of deviation with the corresponding shaping action parameter combination. Its function is to transform the detected geometric or state deviations into executable process adjustment logic.

[0144] The shaping direction can refer to the spatial action orientation taken in response to the current deviation, such as expanding outward along the radial direction of the hole opening to compensate for the small hole diameter, extending downward along the axial direction of the hole depth to compensate for insufficient hole depth, contracting inward along the normal direction of the hole wall to correct the large taper, or tangentially shaping along the edge of the hole opening to eliminate local collapse.

[0145] The shaping sequence can refer to the priority order set to avoid mutual interference when multiple deviations coexist. For example, first correct the defects at the edge of the hole and then adjust the hole depth, or first control the transition trend of the hole wall and then optimize the roundness of the hole.

[0146] Correction intensity can refer to the quantitative representation of the magnitude of a certain correction action, such as the percentage of energy increment, the amount of scanning radius adjustment, the amount of focus defocus, the single dwell time extension value, or the number of repeated corrections.

[0147] In this embodiment, the process of determining the shaping direction, shaping sequence, and shaping intensity is based on the deviations obtained by comparing the actual hole type information extracted in step S6 with the target hole type digital template item by item, and then looking up the corresponding shaping rule in the table to output the result. This result directly determines the adjustment benchmark of each controllable parameter in the subsequent fine finishing and shaping process, and is the key intermediate variable for realizing one-hole-one-policy closed-loop control.

[0148] This application determines the shaping direction, shaping sequence, and shaping intensity of the current hole position by means of table matching based on shaping rules;

[0149] This application determines the shaping direction, shaping sequence, and shaping intensity of the current hole position by weighting the shaping rules and the compensation effect feedback data of similar historical deviations. Furthermore, this application determines the shaping direction, shaping sequence, and shaping intensity of the current hole position by dynamically correcting the shaping rules in combination with the thermal accumulation state of the processing sub-region to which the current hole position belongs and the compensation call record of the previous hole position.

[0150] This application obtains a structured compensation command for driving fine finishing and shaping based on any of the above methods, ensuring that the compensation control amount strictly corresponds to the actual deviation characteristics of the current hole position.

[0151] For example, in this application: when the comparison results show that the current hole position has two deviations, namely, the hole diameter is 0.8 μm too small and there is a local defect at the edge of the hole, the control unit first calls the correction rule corresponding to the small hole diameter, determines that the correction direction is radial outward, the correction order is first priority, and the correction intensity is the scanning radius increased by 1.2 μm; then calls the correction rule corresponding to the defect at the edge of the hole, determines that the correction direction is tangential scanning along the edge, the correction order is second priority, and the correction intensity is the dwell time increased by 2 times; finally, it merges and generates a composite compensation control quantity that includes the radius compensation amount +1.2 μm, the ring scan path is enabled, and the dwell time +2, and writes it to the dedicated storage address of the hole position.

[0152] Step 2: The compensation control quantity is applied to at least one of the following: processing energy distribution, scanning path composition, focus position recall, dwell order, and shaping repetition strategy, and is stored and recalled independently according to hole position.

[0153] Among them, the processing energy distribution can refer to the adjustment of single pulse energy, total pulse train energy, or energy spatial distribution, such as reducing the energy density in the central region to suppress overheating at the bottom of the hole, or increasing the energy in the edge region to compensate for insufficient hole diameter expansion.

[0154] The scanning path can refer to the trajectory of the laser beam within the aperture area. For example, a circular trajectory is used for aperture trimming, a spiral trajectory is used for taper control, and a segmented concentric circle trajectory is used for multi-layer transition trimming.

[0155] Focus position adjustment can refer to setting the position of the optical focusing plane in the Z-axis direction. For example, moving the focus upward to form positive defocus to expand the effective area, and moving the focus downward to form negative defocus to enhance the energy concentration at the bottom of the aperture.

[0156] The dwell order can refer to the arrangement of energy dwell time in different sections of the same scan path, such as extending the dwell time in the edge section of the aperture and shortening the dwell time in the middle section of the aperture wall;

[0157] The repetitive reshaping strategy can refer to the frequency planning of the same reshaping action, such as using two repeated circular scans for areas with severe edge defects, and performing only one tapering reshaping for areas with slight taper deviation.

[0158] In this embodiment, the above five types of objects are all optional output dimensions of the compensation control quantity. The specific combination of their activation is determined by the deviation type of the current hole position and the shaping rule. All compensation control quantities are written into the non-volatile memory with the hole position number as the index key value. When the subsequent fine finishing and shaping process is started, the control unit reads, parses and loads them into the processing execution module according to the hole position number.

[0159] This application, for example, generates a dedicated G-code trajectory instruction for the corresponding hole position based on the scanning path configuration and dwell sequence specified in the compensation control quantity, and calls the instruction to complete the annular trimming of the hole opening during the finishing stage.

[0160] For example, this application dynamically adjusts the Z-axis position of the objective lens and the laser power output according to the focal position specified in the compensation control quantity and the processing energy allocation, forming a low-damage defocusing correction in the transition zone of the hole wall; furthermore, this application inserts two high-dwelling ring scans in the defective section of the hole edge according to the correction repetition strategy and dwell sequence specified in the compensation control quantity, and maintains standard dwell parameters in the remaining sections.

[0161] This application obtains a fine-tuning parameter configuration that precisely matches the current hole position deviation characteristics based on any of the above methods, ensuring the pertinence and reproducibility of the compensation action.

[0162] For example, in this application, for hole number P-047, the compensation control quantity includes the scanning path composition: circular trajectory (diameter = target hole diameter + 1.2μm), dwell order: dwell time of edge section is extended by 30%, and reshaping repetition strategy: executed twice. In the fine finishing and shaping stage, the control system only loads and executes this set of parameters for this hole. Even if the adjacent hole P-048 is located in the same processing sub-area, its compensation control quantity is still read from its own storage address.

[0163] Example 8

[0164] In an optional embodiment, this application further provides that after the compensation control quantity is used for the fine finishing and shaping of the current hole position, it is also written into the control record of the corresponding processing sub-area according to the area index; before processing subsequent hole positions, the layer determination result, termination proximity determination result, hole shape deviation result, and compensation call result of the previous hole position in the same processing sub-area are read, and the compensation control quantity is inherited, modified, or limited according to the positional relationship between hole positions before being used for the fine finishing and shaping of subsequent hole positions, including:

[0165] Step 1: After the compensation control quantity is used for the fine finishing and shaping of the current hole position, it is also written into the control record of the corresponding processing sub-area according to the area index;

[0166] The compensation control quantity (a set of control parameters used to correct the current hole machining process) can be a numerical adjustment command generated in Example 7 that acts on at least one of the following: machining energy distribution, scanning path composition, focus position call, dwell order, and shaping repetition strategy.

[0167] The region index can be a rectangular grid number, a polar coordinate sector number, or a topological adjacency cluster identifier based on the physical coordinates of the optical module substrate. The division criteria include the thermal distribution gradient of the board surface, the difference in copper foil density, the trend of insulation layer thickness variation, or the mechanically clamped deformation sensitive area.

[0168] The processing sub-region can be a local plate area with similar material response characteristics and thermal accumulation behavior, which is defined synchronously when the target hole pattern digital template is established in the S1 stage;

[0169] In this embodiment, the writing action is executed immediately after the current hole position completes the fine finishing and shaping process and the hole wall post-processing is confirmed to be finished. The compensation control quantity, the spatial coordinates of the hole position, the processing sub-area number, the processing timestamp, the layer determination result, the termination proximity determination result, and the hole shape deviation item are packaged together into a structured record and stored in the local cache database or the distributed control record table for subsequent calls by hole positions in the same sub-area.

[0170] Step 2: Before processing subsequent holes, read the layer determination results, termination proximity determination results, hole shape deviation results and compensation call results of the previous holes in the same processing sub-area, and then inherit, correct or limit the compensation control amount according to the positional relationship between holes before using it for the fine finishing and shaping of subsequent holes.

[0171] The layer determination result can refer to the material region identified by at least two of the following during the pre-forming rough processing or interlayer transition processing of the current hole position: reflection response signal, plasma radiation signal, acoustic emission signal, local thermal response signal and coaxial image signal, including the main area of ​​the insulating layer, the interlayer interface transition area or the area near the target termination bottom layer.

[0172] The termination proximity determination result can refer to the remaining bottom thickness range or proximity quantification value output after combining at least two of the reflection response change, thermal response change and acoustic emission change, which is used to characterize the remaining processing amount status of the current hole position from the target hole depth endpoint.

[0173] The hole shape deviation result can refer to at least one of the following: hole profile deviation, hole depth estimation deviation, hole wall transition trend deviation, and hole edge integrity deviation obtained by comparing the actual hole shape information with the target hole shape digital template;

[0174] The compensation call result can refer to the compensation control amount and its effective path actually called by the preceding hole position in the fine finishing process, including the energy reduction magnitude, scanning radius shrinkage amount, focal defocus amount, dwell time increment or number of repeated finishing.

[0175] The positional relationship between holes can be one or more combinations of Euclidean distance, Manhattan distance, topological adjacency or projective overlap, used to quantify the coupling strength of adjacent holes within the range of heat conduction path, stress transfer path or material removal disturbance.

[0176] In this embodiment, for a new hole to be processed, the control system first matches its processing sub-region according to its spatial coordinates, and then extracts the above four results from the control records of the most recent N (N≥1) completed previous holes from the sub-region. When the Euclidean distance between the new hole and a previous hole is less than the preset neighborhood radius (e.g., 200μm) and both belong to the same thermal response sensitive area, the inheritance mechanism is activated—the compensation control amount of the previous hole is directly reused as the initial value. When the new hole is in the overlapping area of ​​the influence of multiple previous holes, the correction mechanism is activated—the compensation control amount of each previous hole is weighted by distance, and the directional offset is adjusted in combination with the relative thermal accumulation level of the current hole in the sub-region. When the new hole is located at the edge of the sub-region or the compensation amount of the previous hole is detected to exceed the process safety threshold, the limiting mechanism is activated—the inherited / corrected compensation control amount is constrained within the preset upper and lower limits to avoid overcompensation leading to hole shape distortion.

[0177] This application may, for example, determine whether to inherit the compensation control amount of the previous hole position based on a joint judgment of the Euclidean distance between hole positions and the thermal response level; this application may also generate the initial compensation control amount of the current hole position by weighted fusion of the compensation call results of multiple previous hole positions and spatial weight coefficients; further, this application may also generate the final compensation control amount for fine finishing and shaping by applying a directional offset or proportional scaling to the inherited compensation control amount based on the thermal accumulation evaluation level of the current hole position in the processing sub-region. Based on any of the above methods, this application obtains feedforward suppression capability for regional system deviations, enabling subsequent hole positions to have an adaptive processing basis before undergoing a complete online sensing and closed-loop correction process, thereby improving the hole depth consistency, hole shape stability, and processing robustness of the entire board micro blind hole array.

[0178] For example, in this application, when performing micro-blind via array processing on a multilayer high-density circuit board for an optical module, the board surface is divided into 16 processing sub-regions of 4×4. When the first via in the 3rd row and 2nd column sub-region is finished and shaped, its compensation control amount (including energy reduction of 5%, scanning radius reduction of 8μm, and focus shift of 3μm) is written into the control record of that sub-region along with the layer position determination result (entered the interlayer interface transition zone), the termination proximity determination result (remaining bottom thickness of 12μm), the via shape deviation result (via diameter 6μm larger), and the compensation call result (actually performed 2 rounds of annular trimming). Before reaching the second borehole in the same sub-region (180 μm away from the first borehole), the system reads the record and determines that the two are in a strong thermal coupling neighborhood. It then inherits the energy and scanning radius compensation amount of the first borehole and applies a -2% reverse fine adjustment to the compensation amount based on the current thermal accumulation level of the second borehole (one level higher than the first borehole). Finally, an initial compensation control amount with a 3% reduction in energy and an 8 μm reduction in scanning radius is generated for its fine finishing and shaping. This process does not require waiting for the second borehole to complete the layer identification and deviation comparison. It can start the adaptation processing in advance, significantly shortening the single-hole closed-loop cycle and suppressing the borehole diameter expansion drift caused by local temperature rise.

[0179] Example 9

[0180] In another embodiment, this application also provides micro-environment linkage control for the current hole machining area during low-damage interlayer transition machining and finishing machining, including:

[0181] Step 1: During the low-damage processing state of interlayer transition and the fine finishing process, micro-environment linkage control is also implemented in the current hole processing area;

[0182] Among them, microenvironment linkage control (directional gas purging, local negative pressure extraction, micro-mist cooling and liquid film slag suppression) can refer to a type of process auxiliary control means that actively regulates the local physical environment of the hole position in the key stage of micro blind hole processing to synergistically inhibit molten material redeposition, reduce heat accumulation, improve slag discharge path and enhance the stability of the hole edge.

[0183] Directional gas purging can be an inert or clean compressed airflow that applies a controllable direction and pressure to the processing area to disperse splashed molten particles, dilute the plasma plume concentration, and cool surface materials.

[0184] Local negative pressure extraction can refer to setting up suction ports around the orifice and maintaining local negative pressure to capture suspended particles, dust and volatile pyrolysis products, and prevent them from falling back and adhering to the orifice wall or orifice edge.

[0185] Micro-mist cooling can refer to the process of ultrasonically atomizing or spraying deionized water or special coolant to form a cloud of submicron-sized droplets, creating a transient endothermic phase change environment in the near field of the orifice, which is used to suppress thermal collapse at the orifice and thickening of the recast layer.

[0186] Liquid film slag suppression can refer to the pre- or simultaneous formation of a continuous or discontinuous liquid film of controllable thickness on the surface to be processed, which is used to absorb laser energy reflection, buffer the impact of molten material splashing, and inhibit the re-adhesion of carbonized resin.

[0187] In this embodiment, the microenvironment linkage control does not operate independently, but is deeply coupled with the processing state: in the low-damage processing stage of interlayer transition, the focus is on suppressing interface tearing and bottom layer overheating; in the finishing and shaping processing stage, the focus is on ensuring the integrity of the orifice edge and the cleanliness of the orifice wall; both require dynamic adjustment of the microenvironment execution parameters based on the specific processing stage of the current orifice position, the real-time acquired orifice contour state, and the predicted / observed direction of residue migration, so as to achieve precise guidance of processing by-products and effective constraint of thermodynamic boundaries.

[0188] This application, for example, uses directional gas purging to control the migration path of splashed particles; it also uses local negative pressure extraction to capture suspended particles and pyrolysis gases; further, it uses the synergistic effect of micro-mist cooling and liquid film slag suppression to inhibit orifice thermal collapse and recast layer formation. Based on any of the above methods, this application achieves the ability to actively intervene in the local physical environment of the processing area, thereby supporting the achievement of low-damage goals during interlayer transition and finishing stages.

[0189] For example, in the low-damage processing stage of interlayer transition, the control system identifies that the current hole position has entered the bottom layer area close to the target termination zone. At the same time, the coaxial image signal shows that the edge of the hole is showing a slight softening trend, and the direction of the change in the reflection signal indicates that the molten material is splashing outward along the upper edge of the hole. At this time, the system immediately starts directional gas purging and adjusts the purging direction from the default radial direction to an upward 30° angle, so that the airflow is in line with the tangent direction of the outer edge of the hole, and guides the splashed material away from the slag collection tank of the hole area. Simultaneously, the local negative pressure extraction in the outer ring area of ​​the hole is started, and the suction port is set 1.2 mm downstream of the main splash path. Within 50 ms after the purging is started, micro-mist cooling is triggered, so that the droplets form a transient cooling cloud at a height of 0.3 mm above the hole, which suppresses edge thermal softening. After transitioning to the fine finishing and shaping stage, the system determined that there was a local expansion in the upper left quadrant based on the newly acquired orifice contour information. Therefore, the purging direction was switched to the upper left 45° angle, and the extraction position was simultaneously shifted to 0.8mm outside the quadrant. Micro-mist cooling was introduced 20ms before the start of the first fine finishing scan to enhance the thermal suppression effect of this weak area.

[0190] Step 2: Microenvironment linkage control includes at least one of directional gas purging, local negative pressure extraction, micro-mist cooling, and liquid film sludge suppression;

[0191] Among them, at least one can refer to allowing only directional gas purging, or only local negative pressure extraction, or only micro-mist cooling, or only liquid film slag suppression, provided that the process effectiveness is met. It can also be combined in pairs (such as directional gas purging + local negative pressure extraction), three combinations (such as directional gas purging + micro-mist cooling + liquid film slag suppression), or all four can be activated. Each combination does not change the essence of this technology. The selection is based on the material composition, thermal sensitivity level, wiring density, and historical processing feedback data of the substrate area where the current hole is located.

[0192] In this embodiment, when the processing object is an insulating layer containing a high proportion of benzoxazine resin, a combination of micro-mist cooling and liquid film slag suppression can be selected to suppress high-temperature cracking and carbonization; when the processing object is a metal adjacent area with high copper content and fast heat dissipation, a combination of directional gas purging and local negative pressure extraction can be selected to improve slag removal efficiency; when the orifice contour detection shows a roundness deviation > ±2μm, directional gas purging is activated and dynamically calibrated in conjunction with extraction position.

[0193] Step 3: Based on the current machining stage of the hole, the state of the hole outline, and the direction of residue migration, switch at least one of the following: purging direction, extraction position, cooling intervention sequence, or liquid film coverage method.

[0194] The processing stage can refer to the current process sub-state, including but not limited to: the start of low-damage interlayer transition processing, the middle stage of interlayer transition, the end stage of interlayer transition, the initial scan of fine finishing and shaping, the intermediate shaping of fine finishing and shaping, and the final closing of fine finishing and shaping; different stages correspond to different heat input intensities, material removal rates and interface disturbance degrees.

[0195] The orifice contour state can refer to the geometric features of the orifice extracted through coaxial image signals, including orifice diameter, roundness, edge sharpness, local collapse amount, burr protrusion length, and edge brightness gradient distribution; this state serves as a direct feedback basis for microenvironment regulation.

[0196] The direction of residue migration can refer to the main motion vector direction of the melt, debris, or dust, which is estimated based on the attenuation trend of reflected signals, the spatial distribution of plasma radiation, the location results of acoustic emission sources, or the tracking of multi-frame coaxial image sequences; this direction is used to predetermine the spatial coordination relationship between purging and extraction.

[0197] In this embodiment, the switching action has a clear cause-and-effect logic: when the system determines that it is currently in the initial scanning stage of fine-tuning and shaping, and the aperture contour status shows that the brightness gradient of the lower right quadrant edge decreases by more than 15% (indicating local softening), and the residual migration direction vector points to the lower right 45°±5°, then the purging direction is automatically switched from the reference 0° to the lower right 45°, the extraction position is switched from the reference annular uniform distribution to fixed-point extraction 1.0mm outside the lower right quadrant, the cooling intervention sequence is advanced from the middle of the scan to 30ms before the start of the scan, and the liquid film coverage method is switched from continuous coverage of the entire aperture area to local thickening coverage of the lower right quadrant (the liquid film thickness is increased by 20%).

[0198] This application, for example, switches the purging direction based on changes in the processing stage; this application, for example, switches the extraction position based on changes in the orifice contour state; further, this application switches the cooling intervention timing based on changes in the migration direction of residues. Based on any of the above methods, this application obtains a strong coupling response capability between the microenvironment execution parameters and the processing dynamic state, thereby ensuring a stable low-damage processing effect under different orifice positions, different regions, and different material conditions.

[0199] For example, in this application: a hole in an optical module carrier is located in a dense area of ​​high-copper wiring. The stacked structure shows that its bottom is a 12μm thick copper conductor layer, the target hole depth is 65μm, and the main body of the insulating layer is FR-4 modified resin. During the low-damage processing stage of interlayer transition, the control system determines that it has entered the interlayer interface transition zone based on the sudden change in the slope of the reflection signal and the shortening of the thermal response delay. At the same time, the coaxial image identifies that the brightness gradient at the top edge of the hole has decreased by 18%, and the analysis of the migration direction of the residue shows that 72% of the molten material splashes along the normal direction at the top of the hole. The system then performs three switches: the blowing direction is switched from horizontal radial to upward 15°, the extraction position is switched from a 3mm ring around the hole to a single point extraction 0.5mm directly above the top, and the cooling intervention sequence is advanced from the original second scan cycle to 40ms before the first scan cycle. This coordinated control reduces the hot collapse at the top edge by 63% and reduces the recast layer thickness to ≤0.8μm, meeting the requirement of the subsequent copper plating process for a hole wall roughness Ra<0.5μm.

[0200] Example 10

[0201] In another embodiment, this application also provides post-processing corresponding to the quality state of the hole wall after the finishing and shaping process, including:

[0202] Step 1: First, inspect the hole wall quality at the current hole location, and select the corresponding post-processing path from among low-temperature plasma cleaning, mild chemical decontamination, microfluidic rinsing, ultrasonic-assisted cleaning, and surface activation treatment based on the inspection results;

[0203] Among them, pore wall quality detection can refer to qualitative or semi-quantitative information characterizing the surface state of the pore wall obtained based on coaxial image acquisition, reflectance spectroscopy analysis, or local thermal response feature extraction; its technical focus is to obtain at least one of the following: pore wall recast layer thickness, molten particle adhesion density, resin carbonization degree, microcrack distribution, and surface roughness level; in this embodiment, the detection is used to provide input basis for the post-processing path, and its detection results directly determine whether to use strong cleaning methods, whether to skip intermediate processing steps, and whether to extend the processing time of a certain type of processing, thereby avoiding insufficient cleaning or over-processing caused by a uniform mode.

[0204] Low-temperature plasma cleaning utilizes reactive particles (such as those generated by non-equilibrium plasma) This method involves oxidative stripping and surface activation of organic contaminants, recast layers, and weakly bonded carbides on the pore walls. In this embodiment, it is used to address situations where pore wall testing results show obvious resin carbonization, recast layer residue, or organic modified layer coverage. It achieves non-destructive decontamination through controlled energy injection, while simultaneously improving the bonding strength of the subsequent metallization interface.

[0205] Mild chemical decontamination involves using a low-concentration, weakly acidic, or complexing solution to briefly immerse or spray the pore walls to dissolve metal oxides, remove trace ion residues, and loosely attached particles. In this embodiment, it is used to address situations where pore wall testing results show mild metal oxidation, ion contamination, or localized particle attachment without significant carbonization, thus avoiding strong corrosive agents from eroding the insulation layer.

[0206] Microfluidic flushing is a physical flushing method that uses microscale channels to guide deionized water or buffer solution to form a directional jet, thereby applying controllable pressure, flow rate, and angle to the orifice and depth. In this embodiment, it is used to address situations where the orifice wall test results show the presence of loose slag, incompletely detached microparticles, or liquid film residues, achieving mechanical removal without chemical intervention by utilizing fluid shear force.

[0207] Ultrasonic-assisted cleaning places the plate to be processed in an ultrasonic tank containing cleaning fluid, and uses the ultrasonic cavitation effect in the 20–100kHz frequency band to enhance the cleaning penetration ability of the microstructure area of ​​the pore wall. In this embodiment, it is used to deal with situations where the pore wall detection results show the presence of embedded microparticles, deposits in the pore wall depression area, or complex morphological areas that are difficult to cover by conventional rinsing, and enhances the deep cleaning effect through cavitation microjets.

[0208] Surface activation treatment involves using low-voltage glow discharge, ozone aeration, or ultraviolet irradiation of a specific wavelength to introduce active functional groups such as hydroxyl and carboxyl groups onto the surface of the pore wall, thereby enhancing its hydrophilicity and the activity of subsequent copper plating reaction. In this embodiment, it is used to address situations where the pore wall test results show that the cleanliness meets the standards, there is no obvious contamination, and the morphology is intact. As a final surface state optimization step, it does not participate in the removal of contaminants, but only enhances the interfacial reaction performance.

[0209] This application, for example, selects whether to enable low-temperature plasma cleaning based on whether the recast layer thickness exceeds a preset threshold in the hole wall quality inspection results; it also selects whether to use microfluidic flushing instead of ultrasonic-assisted cleaning based on whether the microparticle adhesion density is in the low to medium range in the hole wall quality inspection results; further, it directly calls surface activation treatment as the sole post-processing path based on whether the surface roughness Ra value is lower than a set benchmark in the hole wall quality inspection results. This application obtains a post-processing path matching the actual quality state of the current hole location based on any of the above methods, ensuring a balance between thorough cleaning, process economy, and material compatibility.

[0210] For example, in this application: after completing the fine finishing and shaping of the 37th micro-blind hole on a certain optical module carrier, the control system calls a coaxial image sensor to image the hole wall and, combined with reflectance spectral analysis, identifies that the hole wall has a moderate degree of resin carbonization (corresponding to an enhanced spectral absorption peak near 2150 cm⁻¹) and a local recast layer thickness of 1.8 μm; based on this, it is determined that a powerful cleaning path needs to be activated, and the system automatically selects low-temperature plasma cleaning as the main processing method, and superimposes 30s microfluidic rinsing to synergistically remove loose deposits; after the processing is completed, image re-inspection confirms that the carbonization layer is eliminated and the recast layer thickness is reduced to below 0.3 μm, meeting the window requirements of the subsequent copper plating process.

[0211] Step 2: After post-processing, the hole wall quality result of this hole location is written into the subsequent compensation control record of similar holes locations or holes in the same area.

[0212] The hole wall quality result can refer to structured data items formed after testing and post-processing verification, including but not limited to: measured value of recast layer thickness, particle adhesion density level, carbonization degree score, surface roughness Ra value, activated functional group coverage rate, and post-processing path execution identifier; its technical purpose is to quantify and archive the post-processing effect of a single hole, forming the data basis for cross-hole quality feedback; in this embodiment, the result is not stored separately, but is written together with the compensation control amount corresponding to the hole position into the control record table of the same processing sub-area, so that subsequent adjacent holes can read it before fine finishing and use it for parameter inheritance or correction.

[0213] Similar hole positions can refer to other hole positions marked in the design documents as having the same functional area, the same hole diameter specification, the same target conductive layer, and the same wiring density level; hole positions in the same area can refer to other hole positions to be processed that are located in the same processing sub-area in physical space (e.g., a board surface grid divided into 5mm×5mm units); together, they constitute the logical index range for writing back quality data, ensuring that the feedback has process relevance and spatial proximity.

[0214] The subsequent compensation control record is a structured table stored in the local database of the control unit or in the edge computing node. Its fields include at least: hole position number, processing sub-region ID, layer identification result, termination proximity judgment result, hole shape deviation item, compensation control amount, post-processing path identifier, and hole wall quality result. Its technical purpose is to serve as a unified carrier for all process data required for closed-loop optimization. In this embodiment, the record is actively read by the subsequent hole position in step S6 for hole-by-hole compensation control based on the target hole shape digital template. It is used to limit, offset, or weight the energy compensation amount, scanning radius compensation amount, or residence time compensation amount, thereby realizing the full-link coordination of front-end processing, back-end verification, and reverse adjustment.

[0215] For example, based on the measured recast layer thickness of hole position 37 being 15% higher than the average for similar holes, the control system automatically reduces the energy compensation of hole position 38 by 8% before it enters the finishing process to suppress the tendency of recast layer formation. Another example is based on the high particle adhesion density of hole position 37, which prompts the system to increase the timing of micro-mist cooling intervention and initiate local negative pressure extraction before processing hole position 39 to improve slag removal conditions. Furthermore, based on the surface activated functional group coverage rate being lower than the threshold in hole position 37, a short-pulse focal plane disturbance operation is added during the finishing stage of hole position 40 to enhance the micro-roughness and active point density of the hole wall. This application achieves dynamic adaptation capability to the processing parameters of subsequent holes based on any of the above methods, enabling compensation control to not only respond to current hole deviations but also integrate historical quality feedback, improving overall plate consistency.

[0216] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for fabricating high-precision micro blind holes for optical modules, characterized in that, include: S1. Obtain the stacked structure information, target conductive layer position and micro-blind hole design requirements of the optical module substrate to be processed, and based on the stacked structure information, target conductive layer position and micro-blind hole design requirements, establish a target hole digital template corresponding to the hole position to be processed. The target hole digital template is used to characterize the target hole depth, hole opening forming boundary and hole bottom forming requirements. S2. Based on the target hole pattern digital template, perform pre-forming rough machining on the hole to be processed, and simultaneously collect online feature signals characterizing the material removal state during the pre-forming rough machining process; S3. Identify the material layer position of the current hole position based on the online feature signal, and determine the remaining bottom thickness of the current hole position relative to the target hole depth; when it is identified that the current hole position has entered the target termination area, switch the processing state of the current hole position from the pre-forming rough processing state to the interlayer transition low-damage processing state. S4. Extract the actual hole shape information of the current hole position, and compare the actual hole shape information with the target hole shape digital template to obtain the hole shape deviation of the current hole position, and generate a compensation control quantity for correcting the current hole position processing based on the hole shape deviation. S5. Perform fine finishing and shaping on the current hole position according to the compensation control amount to complete the micro blind hole termination depth control and hole shape correction, and perform post-processing on the hole wall corresponding to the hole wall quality state after the fine finishing and shaping is completed.

2. The method according to claim 1, characterized in that, The process of establishing a target hole pattern digital template corresponding to the hole position to be processed includes: Read the stack-up structure file corresponding to the hole location, extract the insulation layer thickness, target conductive layer location, interlayer interface location, and wiring boundary information around the hole location, and create a template index according to the hole location; The target hole pattern digital template includes at least three of the following: target hole depth, hole opening contour boundary, hole bottom retention area, and hole wall transition constraint, to form a hole pattern description set corresponding to a single hole location.

3. The method according to claim 2, characterized in that, The online feature signals include at least two of the following: reflection response signals, plasma radiation signals, acoustic emission signals, local thermal response signals, and coaxial image signals. During the pre-forming roughing process, the online feature signals are collected according to the processing pulse segment or scanning circle, and at least two of the following are extracted: signal amplitude, change slope, abrupt change position, thermal decay sequence and hole edge information, as the removal status input of the current hole position.

4. The method according to claim 3, characterized in that, The step of identifying the material layer where the current hole is being processed based on the online feature signal includes: Pre-establish characteristic response rules for the main insulating layer region, the interlayer interface transition region, and the bottom layer region near the target termination; During the processing, the currently acquired online feature signals are matched with the feature response rules, and the material layer of the current hole position is determined by combining the previous hole processing records of the area to which the current hole position belongs.

5. The method according to claim 4, characterized in that, The determination of the remaining bottom thickness of the current hole position relative to the target hole depth includes: At least two of the changes in reflection response, thermal response, and acoustic emission are combined to generate a termination proximity determination result. When the termination proximity determination result enters the preset switching interval, the processing trajectory and processing mode corresponding to the pre-forming rough processing are stopped, and the processing trajectory, action path and pulse scheduling mode corresponding to the interlayer transition low damage processing state are called instead.

6. The method according to claim 5, characterized in that, The extraction of the actual hole type information at the current hole position includes: Obtain at least two of the following information for the current hole location: hole profile, hole depth estimation, hole wall transition trend, and hole edge integrity. The acquired information is then matched with the hole outline boundary, hole depth termination position, hole wall transition constraint, and hole edge constraint in the target hole type digital template to obtain the deviation of the current hole position in the hole opening, hole depth, hole wall, and edge items.

7. The method according to claim 6, characterized in that, The step of generating a compensation control quantity based on the hole shape deviation to correct the current hole position machining process includes: According to the shaping rules corresponding to the deviation, determine the shaping direction, shaping sequence and shaping intensity of the current hole position, and generate the compensation control quantity corresponding to the current hole position; The compensation control quantity acts on at least one of the following: processing energy distribution, scanning path composition, focus position calling, dwell order, and shaping repetition strategy, and is stored and called independently according to hole position.

8. The method according to claim 7, characterized in that, After the compensation control quantity is used for the fine finishing and shaping of the current hole position, it is also written into the control record of the corresponding processing sub-area according to the area index. Before processing subsequent holes, the layer determination result, termination proximity determination result, hole shape deviation result and compensation call result of the previous holes in the same processing sub-area are read. The compensation control quantity is inherited, modified or limited according to the positional relationship between holes before being used for the fine finishing and shaping of subsequent holes.

9. The method according to claim 1, characterized in that, During the low-damage processing state of interlayer transition and the fine finishing process, micro-environment linkage control is also implemented in the current hole processing area; The microenvironment linkage control includes at least one of directional gas purging, local negative pressure extraction, micro-mist cooling, and liquid film sludge suppression. Based on the current processing stage of the hole, the state of the hole profile, and the migration direction of the residue, switch at least one of the following: purging direction, extraction position, cooling intervention sequence, or liquid film coverage method.

10. The method according to claim 1, characterized in that, The post-processing performed on the hole wall after the finishing and shaping process, corresponding to the hole wall quality state, includes: First, the hole wall quality of the current hole location is inspected, and the corresponding post-processing path is selected from low-temperature plasma cleaning, mild chemical decontamination, microfluidic rinsing, ultrasonic-assisted cleaning and surface activation treatment according to the inspection results; After post-processing, the hole wall quality results for that hole location are written into the subsequent compensation control record for similar holes or holes in the same area.

Images