Digital direct recording device comprising real time analysis and correction of recorded artwork by dividing the artwork into submodules

TWI945182BActive Publication Date: 2026-11-11VISITECH LITHOGRAPHY AS
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
TW111109315
Authority / Receiving Office
TW · TW
Patent Type
Patents
Current Assignee / Owner
Priority Date
2021-03-17
Filing Date
2022-03-15
Publication Date
2026-11-11
Estimated Expiration
2042-03-14

AI Technical Summary

Technical Problem

Existing methods for correcting positional and rotational errors in electronic components during the manufacturing process are time-consuming, limiting the throughput of the production line due to the need for rerouting and rasterizing negatives, which is inefficient and costly.

Method used

A system and method for real-time error correction of predetermined negatives by dividing them into submodules and performing offline preprocessing and online real-time processing, utilizing a substrate recorder with optical elements to project the corrected negatives onto the substrate, minimizing rasterization time and ensuring accurate electrical connections.

Benefits of technology

This approach significantly reduces rasterization time by a factor of 100-10,000 and allows instantaneous error correction during the recording process, improving production yield and throughput without the need for new negatives, thus enhancing manufacturing efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure TWG2TB001919927_001
    Figure TWG2TB001919927_001
  • Figure TWG2TB001919927_002
    Figure TWG2TB001919927_002
  • Figure TWG2TB001919927_003
    Figure TWG2TB001919927_003
Patent Text Reader

Abstract

A method for digitally directly recording a film representing electrical connections of components on a substrate includes the following steps: a) receiving data representing the film; b) analyzing the film representation to identify similar and unique portions and to identify the orientation of components in the film; c) dividing the film into modules corresponding to the identified portions, providing a set of unique modules and a set of redundant modules; d) rasterizing each unique module to provide a rasterized module; e) dividing the rasterized module into sub-modules; f) receiving measurements of the positions of components on the substrate; g) receiving measurements of the positions of the substrate; h) calculating the difference between the measured position of the component and the film position of the component; i) calculating modifications for each of the sub-modules to compensate for the difference; and j) recording the modified sub-modules onto the substrate to form a modified film on the substrate.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to a process for compensating for positioning tolerances introduced when positioning electronic components (e.g., dies) on a substrate during processes such as fan-out panel-level package (FOPLP) or fan-out wafer-level package (FOWPL). When electronic components are positioned, for example, by a pick-and-place machine, there may be offset or rotation compared to their nominal positions. The film describing the electrical connections of the electronic components must be corrected accordingly to provide correct electrical connections. This process is preferably performed immediately to avoid limiting the output of production lines containing one or more electronic modules. This invention helps improve yield during the manufacturing process because fewer electronic components have to be rejected due to poor positioning and therefore lack of electrical connections, while avoiding the disadvantage of having to generate new rasterized films. Prior Technology

[0002] The electronic components industry uses various technologies to achieve electrical connections between electronic components or electronic component modules. In one embodiment, a substrate serves as a temporary carrier, and adhesive tape is used to place components in predetermined positions on the substrate. For example, individual electronic components can be positioned using a "pick-and-place" machine; this process introduces some positional and rotational errors due to the precision limitations of such pick-and-place equipment, which requires high speed and high throughput. The electronic components are then molded with a compound coating, and during this molding process, the electronic components may move or rotate slightly. The substrate and adhesive tape are then removed in a specific process. What is obtained is a compound layer in which the electronic components and their electrical connections / pads are visible from one side.

[0003] Next, a predetermined film is applied to the top of the electronic component, for example, by lightly exposing it to the carrier substrate, to achieve electrical connection between the components. If the positioning of the electronic component does not match the predetermined electrical interconnect film due to tolerances during component placement, it is necessary to measure the actual position of the component and to perform error correction on the film at some locations around the electronic component. Then, the film is applied to the component on the substrate to provide the correct electrical connection.

[0004] Several techniques for reconfiguring film have been identified. US8799845 and US4835704 describe a process in which a new pattern is generated for each substrate with assembled electronic components based on measurements of the actual positions of the electronic components. US8799845 describes a process that allows for routing electrical connection tracks in a new order, while US4835704 describes a process for rerouting portions of the tracks. Both techniques are very time-consuming because the film needs to be rerouted before data can be transferred to a recording device (e.g., a direct imaging machine that applies the pattern to the substrate, where components will be mounted later in the process), and subsequently rasterized. The rerouting and rasterization processes are very time-consuming, limiting the overall production yield. If the electronic components are properly positioned and within a certain range, rerouting is unnecessary; only error correction is performed on the rasterized film.

[0005] To avoid any limitations in production time due to film error correction, it is necessary to pre-process the predetermined film during the offline process. This pre-processing is done in a way that allows for on-the-spot error correction of the original film when it is applied, rather than generating new films for each substrate to be produced.

[0006] In the following description, the term "film" refers to a design in the form of a database describing interconnect structures and via structures used to provide interconnections between electronic components. The term "rasterization" refers to the task of acquiring an image described in a vector graphics format (shape) and converting it into a raster image, which is a series of pixels, dots, or lines that, when displayed together, produce an image represented by shapes. Summary of the Invention

[0007] The object of this invention is to provide a system and method in which error correction of a predetermined film can be performed in real time during the process of recording a predetermined film onto a substrate by means of a substrate recorder. The substrate recorder includes a recording device unit, which includes optical elements, such as a light source or projector for projecting the film onto the substrate. In the following description, the term "light source" will sometimes be used as a general term for optical elements including an exposure system and may be used interchangeably with the term recording device unit or recording unit.

[0008] To meet real-time requirements, the pre-reserved film needs to undergo certain offline preprocessing steps and online real-time processing steps, as shown in Figures 1 and 2. These steps will be described in more detail later. The offline process is typically not performed during the process of recording data onto the substrate, so recording several substrates does not affect the total yield. The offline process is performed before the online process and can be performed once per batch or per group of substrates, with a minimum of one substrate recorded, but in most cases, it includes recording multiple substrates using the same pre-reserved film. The offline preprocessing steps include dividing the pre-reserved film into many modules. These modules may contain one or more levels of sub-modules. Pre-reserved films used in FOPLP / FOWLP applications typically contain many equal modules distributed on top of the substrate. Therefore, all modules in the pre-reserved film can be represented by a list of modules and their respective positions on the pre-reserved film, rather than representing the film as a single large file. Any additional parts of the pre-defined film that are not shown multiple times on the pre-defined film, such as frames or objects shown only once, can also be represented by a module and a list of their individual positions on the pre-defined film module. Therefore, the complete pre-defined film will be represented by a number of unique modules and a list of the positions of each of these unique modules. Since all equivalent modules can be rasterized only once, rather than rasterizing the entire pre-defined film, the rasterization process will be shortened by 100 to 10,000 times, because there can be 100 to 10,000 equivalent modules in the pre-defined film. The rasterization process requires significant CPU resources and a certain processing time, depending on the available CPU resources. Therefore, even in offline processes, the amount of data to be rasterized should be minimized to save processing time, as this will allow for recording prototype batches or groups of substrates with a low number of substrates per batch or group in the shortest possible time.

[0009] Following rasterization, preprocessing involves further dividing the module into one or more sub-modules at one or more levels (Figure 4a). The size and shape of the sub-modules vary depending on the properties of the predetermined film area covered within the sub-module. Such properties may include a description of the allowable scaling (warping) and positioning (rotation and offset) of each sub-module. Each sub-module may be subdivided into smaller sub-modules until the lowest level of the sub-module is reached. The lowest level of the sub-module is referred to as a tile in the following description. Each tile has a set of parameters that reflect the parameters of the sub-module at the level above the tile. The tile serves as the basis for further processing of the predetermined film during offline processing. For example, a portion of the predetermined film may describe an area where electronic components are mounted. Such an area may be contained within a sub-module. The electronic components do not change shape but may be positioned (offset and rotated) differently between substrates depending on the positioning tolerance of the pick-and-place mechanism. Therefore, the film describing the electrical connection points of these components can only be positioned accordingly. This means that if such areas of the predetermined film are scaled, the electrical connections will not fit the components to be connected. Therefore, such areas covered by sub-modules are represented by tiles, which have parameters describing how much positioning must be applied to ensure correct electrical connections. Other sub-modules describe areas of the predetermined film with electrical tracks that connect electronic components together. In these sub-modules, track error correction can be performed during the in-line process to compensate for positioning tolerances of the electronic components on the substrate. Therefore, the tiles in these sub-modules have the property of allowing scaling and positioning, slightly adjusting the track positions to provide correct electrical connections between components (Figure 7).

[0010] To minimize latency during the transfer of the pattern to the substrate recorder, lossless compression (e.g., stroke encoding) can be performed on the pattern before intermediate storage. During this stage, substrate recorder parameters are configured, such as initializing motion drive and recording parameters.

[0011] The online process provides real-time error correction and recording of a pre-defined rasterized film segmented into tiles. Error correction is performed based on the input of measured position coordinates of multiple reference objects (such as regional reference points for each module on the substrate, provided by, for example, an external measuring machine). These position coordinates provide information about how the electronic components to be electrically interconnected are actually positioned on the substrate relative to each other. Furthermore, the measured position coordinates provide information about the substrate's position within the substrate recorder, often referred to as global reference points, used for error correction. This can be accomplished by an algorithm that calculates the required scaling and positioning for each tile to provide accurate error correction for the pre-defined film.

[0012] Depending on the available machine configuration (e.g., the number of recording heads and their internal spacing in the actual substrate recorder) and the dimensions of the actual substrate and film, the recording device (e.g., one or more recording heads) requests information from the tile server to execute the necessary information for recording error-corrected film onto the substrate. The tile server is where tiles are stored intermediately during the offline process. Furthermore, data containing calculated scaling and positioning information for each tile is provided from the online warp calculation to provide error correction for the predetermined film. When the recording device receives all tiles, along with the position and scaling and positioning of each tile, recording of the fully error-corrected film can be enabled. Global error correction and regional error correction can be performed simultaneously. An example of the resulting error-corrected film is depicted in Figure 7, and will be described in more detail below. Simple Explanation of the Diagram

[0013] The invention will now be described in more detail by way of example and with reference to the accompanying drawings.

[0014] Figure 1 illustrates the offline preprocessing steps according to an embodiment of the present invention.

[0015] Figure 2 illustrates the online real-time processing steps according to an embodiment of the present invention.

[0016] Figure 3 illustrates an example of a pre-reserved film used in FOPLP / FOWLP applications.

[0017] Figure 4 illustrates the preprocessing steps for dividing a module into one or more sub-modules at one or more levels.

[0018] Figure 5 shows an example of a list of sub-module locations.

[0019] Figure 6 shows an example of a calculated warpage list describing a submodule.

[0020] Figure 7 shows a negative that has been incorrectly corrected.

[0021] Figure 8 shows an embodiment of another redistribution layer (RDL) for interconnecting or routing electronic components to electrical connections.

[0022] Figure 9 illustrates the dimensional changes of the sub-module.

[0023] Figure 10 illustrates an example of the design rules for film negatives. Implementation

[0024] Figures 1 and 2 illustrate embodiments of the method according to the present invention, with further explanation in the other figures. The process is divided into an offline process shown in Figure 1 and an offline process shown in Figure 2.

[0025] Figure 1 illustrates the entire offline process. This process can be performed once per batch or per set of jobs. For each batch or set of jobs, there is a predetermined film for all substrates to be used for recording a specific film. An example of film 31 is shown in Figure 3. The purpose of dividing the process into offline and online portions is to ensure that processing steps are performed only once per batch or per set of jobs. Due to limited processing capacity, these processing steps are often time-consuming, and therefore need to be performed outside the recording process to avoid limiting overall recording throughput.

[0026] The offline process is typically performed as a preprocessing step before the online processing steps shown in Figure 2.

[0027] The offline preprocessing steps include receiving data representing the film, such as a CAD file, in step (1), and dividing the predetermined film into multiple modules in step (2) before rasterizing the film in step (3). Dividing the predetermined film 31 into modules 32 and 33 is based on analyzing the film to identify similar and unique portions and to identify the orientation of components in the film. The identification of similar or equal modules 33 provides multiple modules, which can be represented by a unique module and a list containing the positions of each of the equal and redundant modules in the predetermined film 31. The non-redundant portion 33 of the predetermined film, i.e., the unique module that is only one, is also represented by the module and a list of its positions in the predetermined film 31. Since the unique module is rasterized only once, instead of rasterizing all redundant modules in the entire predetermined film, the rasterization process time will be reduced by 100 to 10,000 times, because there can be 100 to 10,000 equal modules in the predetermined film. Another advantage of dividing a predetermined film into modules is that the open areas between modules do not require rasterization; only the areas containing the film need rasterization. Therefore, the process of generating modules from a predetermined film saves processing time during rasterization. This is desirable even in offline processes, especially when recording film on small prototype batches or sets of recording substrates.

[0028] After the rasterization of non-redundant modules 32 and 33, these modules are divided into one or more levels of sub-modules. Different levels mean that one sub-module can be inside and part of another sub-module. The size of the sub-module can vary and, in some cases, can be wider than the width of the working area of ​​the recording device unit / light source in the substrate recorder.

[0029] This is illustrated in Figure 9, where the width of sub-module 91 is greater than the width 92 that the projection of the recording device unit can cover. Therefore, several recording device units / light sources can be used to record the full width of the sub-module. To achieve this, the substrate recorder may include several light sources.

[0030] To limit data transmission to individual recording units within the baseboard recorder, sub-modules can be represented in a flexible format, allowing for efficient transmission and minimal processing within the recording unit. This can be achieved by dividing each sub-module into a lower-level set of sub-modules, where each lower-level sub-module is smaller than the one above it.

[0031] Sub-modules are illustrated in Figures 4a and 4b. The lowest-level sub-modules 44 and 411 containing the actual rasterized bitmap data are called tile, with a minimum size of 2x2 pixels. This sub-module (tile) serves as a building block for further processing in both offline and online processes. Each tile represents one or more defined finite regions 43 of the entire predetermined film, having electrical connections (e.g., pads or vias) to electronic components (e.g., transistors), wherein the internal relative distances and positioning of the electrical connections are typically offset or rotated only slightly, or if so, scaled or deformed only to accommodate the relative displacement between the components and the surrounding layout.

[0032] Figure 8 illustrates an example where the brick is part of the ReDistribution Layer (RDL) 81 in the FOPLP / FOWLP application.

[0033] In this configuration, the tile can be defined as a limited area of ​​the entire predetermined substrate of the module, where all or some of the electronic components are interconnected or routed to another electrical connection via vias 85 in a redistribution layer (RDL). The electronic components (e.g., dies 83) can then be electrically connected to the substrate 82 using solder balls 84 facing the upper RDL. This RDL stack allows all electrical signals from the dies to be connected to the substrate while maintaining a small overall package size. In this configuration, the internal relative distances and positioning of the electrical connections on the individual RDLs are typically only offset and / or rotated, and if so, only slightly scaled or deformed to fit the other RDLs.

[0034] The dimensions of tiles 44 and 411 are square or rectangular, depending on the optimal way to cover the area of ​​the higher-level sub-module. Tiles may also have other suitable shapes to optimize the adaptation to the covered area. Furthermore, the size and shape will depend on the content of the defined data protocol that allows image data to be transmitted to the recording device. In addition, tiles inherit information from higher-level sub-modules, namely, how they are allowed to be reshaped in the output on the recording device (i.e., during recording).

[0035] After defining the tiles, the offline preprocessing step (5) of Figure 1 includes data compression for each tile. Data compression of the tiles can be performed using a lossless compression algorithm to achieve minimal transmission time during the subsequent online process of Figure 2. Additionally, a list of the locations of each tile is generated. An example of this list is shown in Figure 5.

[0036] The final step (6) of the offline process shown in Figure 1 involves configuring the substrate recorder to prepare general recording parameters, such as setting motion drive and recording energy parameters. This can be done during the offline process and before the work of each batch or group, where each batch / group includes a number of substrates to be recorded.

[0037] The online process shown in Figure 2 is initiated by step (1), which involves inputting the measured position coordinates of multiple reference objects, such as regional reference points provided by, for example, an external measuring machine. These position coordinates provide information on how the electronic components to be electrically interconnected are positioned relative to each other on the substrate. Furthermore, in step (2), the measured position coordinates provide information about how the actual substrate is positioned within the substrate recorder, often referred to as global reference points, which is passed as input to step (3). Step (3) involves calculating the error correction parameters or warpage dataset for each sub-module.

[0038] By using these positional measurements, both area correction and global correction can be performed on the film simultaneously. This is because the global measurement position coordinates from the recorder are used to recalculate the area correction relative to the substrate recorder coordinate system. Such correction calculations are typically performed using bilinear, spline, or similar suitable interpolation methods. The recalculated area correction is then used to calculate the offset and deformation of each tile within its individual sub-module. As the tiles and corresponding lists 52 describing the warpage of each tile are transmitted to the respective recording units of the substrate recorder, the internal relative positions of the individual recording units are also compensated to produce a complete film without overlap or gaps between the recording results of the various recording units within the substrate recorder.

[0039] In step (4) of the online process, a sequence of tiles to be recorded and assigned to recording units is generated. The sequence is generated based on available configurations, such as the number of recording heads and their internal spacing in the actual substrate recorder, and the dimensions of the actual substrate and film. The sequence includes: a) Record the error-corrected film onto the substrate as required; b) A list describing the calculated warpage of the tiles (represented by the x and y offsets of each corner of each tile). Figure 6 illustrates an example of such a list.

[0040] Once the complete sequence of tiles is ready, the tiles and the tile warpage information list are transmitted to the substrate recorder. To obtain immediate requirements with minimal delay before recording in the online process, the tiles are transmitted in the order they were recorded in step (5). This is commonly referred to as the streaming process. Since each tile is compressed during the offline process (Figure 1, step (5)) to save transmission time during tile streaming, it will be decompressed in step (6) after being received by the substrate recorder. In step (7), the substrate recorder places the tiles in the correct positions, performs warping as described in the tile warpage information list, and records the result on the substrate.

[0041] Figure 7 illustrates the resulting error-corrected recorded film, which includes module 71 with sub-modules 72, 73 and warped tiles 74, 75. The electrical tracks in the predetermined film have been error-corrected compared to the tracks 45 described in Figure 4a, because the individual tiles 44, 411 within the sub-modules have been warped according to a list 62 describing the calculated warpage of each tile to achieve the shape described in 74. In the bitmap domain of the rasterized predetermined film, the coordinates of the angles of each tile 61 are given in pixels or decimal places. Processing steps (3), (4), (5), and (6) described in Figure 2 are performed within milliseconds, representing an instantaneous process, since the recording process in step (7) takes several seconds depending on the properties of the recording substrate. Therefore, the instantaneous error correction during recording according to the present invention represents a significant improvement compared to, for example, US8799845 and US4835704.

[0042] During the calculation of the warp set in step (3), control can be performed in step (8) for a predetermined set of design rules set by the user (design rule check). Examples of such design rules are shown in Figure 10—such design rules may include information such as allowable line tolerance 101 and line spacing tolerance 102 for the film. For each tile, the applied warp (scaling and positioning) described in List 62 of Figure 6 can be checked to see if it conforms to the predetermined design rules. If not, information describing which tiles do not conform to the design rules is communicated to the user in step (9) so that the user can take appropriate action. Such actions may include stopping the recording process, or, if the recording process is allowed, discarding only the recorded areas that are not within the design rules.

[0043] After the recording process step (7), the recording substrate can undergo visual inspection, for example, in an external Automatic Optical Inspection (AOI) machine. This inspection can utilize reference data from the recorder, so the recorder must prepare the data for recording in step (10) in such a way that the AOI machine can utilize this data in step (11).

[0044] 31: Film 32: Module 33: Module 43: Defining a limited area 44: Picture Brick 45: Track 52: List 61: Picture Brick 62: List 71: Module 72: Sub-module 73: Sub-module 74: Warped Bricks 75: Warped Bricks 81:RDL 82:Substrate 83: Grain 84: Welding ball 85: Electrical Through Hole 91: Sub-module 92: Width 101: Line tolerance difference 102: Line spacing tolerance 411: Picture Brick

[0045] Domestic storage information (please note in order of storage institution, date, and number) none Overseas storage information (please note in the order of storage country, institution, date, and number) none

Claims

1. A method for digitally direct recording a film representing multiple electrical connections of multiple components on a substrate, the method comprising the steps of: a) receiving data representing the film; b) analyzing the film representation to identify multiple similar portions and multiple unique portions, and identifying multiple orientations of the components in the film; c) dividing the film into multiple modules corresponding to the identified portions, providing a set of unique modules and a set of redundant modules; d) rasterizing each unique module to provide multiple rasterized modules; e) dividing the rasterized modules into multiple sub-modules; f) receiving multiple measurements representing multiple positions of the components on the substrate; g) receiving multiple measurements representing the positions of the substrate; h) calculating multiple differences between the measured positions of the components and the film positions of the components; i) calculating multiple modifications of each of the sub-modules to compensate for the differences; and j) The modified sub-modules are recorded onto the substrate to form a modified film on the substrate.

2. The method as described in claim 1, comprising a step in which the sub-modules are divided into a plurality of further levels of sub-modules, and the lower-level sub-modules represent smaller regions.

3. The method as described in claim 2, wherein the sub-modules are analyzed to identify multiple sub-parts having multiple specific features.

4. The method as described in claim 3, wherein the features are one or a combination of the following: - multiple regions in which the relative distance between multiple elements in the film must be maintained, - multiple regions in which the relative distance between multiple elements in the film can be scaled or deformed, - multiple regions representing a larger area than that can be recorded in a single operation.

5. The method as described in any one of claims 1 to 4, wherein each redundant module is represented by a unique module and a location, and wherein a list containing a plurality of instructions for modifying each module is generated.

6. The method as described in any one of claims 1 to 4, wherein the modification of the data of the film is performed in accordance with a plurality of predetermined design rules.

7. The method as described in any of claims 1 to 4, wherein the measurement value in step f) is received from a remote measurement station.

8. The method as described in any of claims 1 to 4, wherein the measurements in step g) are received from an internal measurement system.

9. A system for digitally direct recording a film representing multiple electrical connections of multiple components on a substrate, the system comprising a substrate recorder including a recording device unit, and multiple measuring members for measuring multiple positions of the components on the substrate and / or the positions of the substrate, wherein the substrate recorder includes: - multiple memory members adapted to receive and store data representing the film, and - a processing device adapted to - analyze the film representation to identify multiple similar portions and multiple unique portions, and identify multiple orientations of the components in the film, - divide the film into multiple modules corresponding to the identified portions, providing a set of unique modules and a set of redundant modules, - rasterize each unique module to provide multiple rasterized modules, - divide the rasterized modules into multiple sub-modules, - receive multiple measurement values ​​representing multiple positions of the components on the substrate, - receive multiple measurement values ​​representing the positions of the substrate, - Calculate multiple differences between the measured positions of the components and the film positions of the components; - calculate multiple modifications for each of the sub-modules to compensate for the differences; and - send the modified sub-modules to the recording device unit for recording onto the substrate to form a modified film on the substrate.

10. The system as claimed in claim 9, wherein the processing apparatus is further adapted to divide a plurality of sub-modules into a plurality of further levels of sub-modules, and the lower-level sub-modules represent smaller regions.

11. The system as claimed in claim 9 or 10, wherein the processing apparatus is adapted to analyze the sub-modules to identify multiple sub-parts having multiple specific characteristics.

12. The system as described in claim 9 or 10, wherein each redundant module is represented by a unique module and a location, and wherein a list containing a plurality of instructions for modifying each module is generated.

13. The system as described in claim 9 or 10, comprising a remote measurement station for representing multiple measurements at multiple locations of the components on the substrate.

14. The system as claimed in claim 9 or 10, wherein the substrate recorder includes an internal measurement system for performing a plurality of measurements representing the position of the substrate.

Citation Information

Patent Citations

  • Method and computer program product for computational metrology

    TW202040260A

  • Gradient assisted image resampling in micro-lithographic printing

    US20100142838A1

  • Adaptive patterning for panelized packaging

    US20110202896A1