A multi-ic electromagnetic coupling decoupling guidance method based on system-level EMC data assets

Abstract

This invention discloses a multi-IC electromagnetic coupling decoupling guidance method based on system-level EMC data assets, belonging to the field of electromagnetic compatibility design technology. The method includes: acquiring user-specified IC combination information; retrieving a matching baseline coupling coefficient matrix from system-level EMC data assets; calculating the comprehensive coupling contribution value of each IC based on the matrix; identifying the dominant path based on path decomposition and screening key coupling paths; generating decoupling instructions based on the key paths, the instructions including mandatory and reserved instructions, and attaching design parameter reference values, effect predictions, and applicable conditions; and storing the decoupling instructions in the user's local design knowledge base for subsequent retrieval. This invention quantifies the coupling between multiple ICs through data assets, identifies the dominant path, generates executable decoupling instructions, and supports experience iteration, significantly improving the efficiency and accuracy of EMC design. This invention can be integrated into intelligent EMC design instruments or implemented as a standalone software product.

Description

Technical Field

[0001] This invention belongs to the field of electromagnetic compatibility (EMC) design technology, specifically relating to a multi-IC electromagnetic coupling decoupling guidance method based on system-level EMC data assets. It is particularly suitable for integration into EMC intelligent design instruments or as computer software products (including standalone applications, software plug-ins, embedded software, etc.), providing hardware engineers with executable decoupling instructions. Background Technology

[0002] With the increasing integration of electronic systems and the widespread adoption of multi-core architectures, modern electronic systems often contain dozens or even hundreds of integrated circuits (ICs). When these ICs work together in a confined space, the electromagnetic coupling effect between them exacerbates the complexity of the system's electromagnetic environment, becoming a key factor affecting electromagnetic compatibility. Coupling between ICs can propagate through conduction paths (power / ground plane, signal lines) or radiation paths (spatial near-field coupling), leading to a more complex system electromagnetic environment and subsequently causing excessive radiated emissions, decreased receiver sensitivity, or malfunctions.

[0003] Traditional IC layout design relies primarily on engineers' experience and general design rules, which has the following shortcomings:

[0004] First, existing methods lack the ability to quantitatively analyze complex electromagnetic couplings between ICs. Engineers typically lay out their ICs based on principles such as "keeping high-speed ICs away from sensitive circuits" and "separate partitions for power ICs," but they cannot quantitatively assess the coupling strength between different ICs, nor can they determine which coupling paths need to be prioritized. This experience-driven design often leads to over-design or under-design.

[0005] Second, while EMC simulation analysis methods can simulate electromagnetic coupling between ICs, they require the establishment of accurate IC models and complete PCB models, resulting in long modeling cycles, high computational resource consumption, and the accuracy of simulation results being highly dependent on model precision. More importantly, the simulation process itself is not digitized, failing to provide traceable decision-making basis for the design of subsequent similar systems.

[0006] Third, existing technologies struggle to quantitatively decompose the coupling paths between ICs in complex electromagnetic coupling and lack the ability to identify the dominant coupling path. Engineers often only know that "strong coupling exists between ICs," but struggle to distinguish whether the main cause of coupling is power conduction, signal coupling, or spatial radiation. This leads to blind selection of EMC measures; for example, implementing numerous filtering devices for coupling problems dominated by spatial radiation often fails to achieve the desired suppression effect.

[0007] Fourth, existing design methodologies struggle to generate reusable experience assets. Design experiences from different projects and engineers are scattered across personal documents or individual files, making them unsystematic and unreusable. When faced with similar multi-IC systems, designers need to re-analyze the data and cannot quickly access relevant experience from past designs.

[0008] Patent application CN114117991A discloses a method, apparatus, and system for determining the location of decoupling capacitors in a chip. It determines the optimal capacitor location through simulation iteration. However, this method only addresses single-chip decoupling, does not consider coupling between multiple ICs, and does not generate decoupling instructions that include mandatory and reserved measures. Patent application CN115062578A discloses an inter-stage decoupling evaluation method for millimeter-wave integrated circuits. However, this method only evaluates coupling conditions and does not provide specific decoupling guidelines. Patent application CN119227467A discloses an electromagnetic interference suppression method and system for chips. It identifies and suppresses interference through simulation modeling, but it does not quantify inter-IC coupling based on system-level data assets, nor does it output layout optimization rules.

[0009] In summary, existing technologies lack a multi-IC system layout optimization method that can quantify and analyze electromagnetic coupling between ICs, decompose coupling paths, identify dominant paths, generate executable decoupling instructions, and continuously evolve through practice. Summary of the Invention

[0010] The purpose of this invention is to provide a multi-IC electromagnetic coupling decoupling guidance method based on system-level EMC data assets, to solve the problems in existing technologies such as lack of quantitative analysis of inter-IC coupling, difficulty in identifying dominant paths, and fuzzy and ineffective optimization rules. This method can be integrated into intelligent EMC design instruments or implemented as computer software products (including standalone applications, software plugins, etc.), providing engineers with clear and quantifiable decoupling instructions, reducing reliance on expert experience, and minimizing design result discrepancies caused by human factors.

[0011] To achieve the above objectives, the present invention provides the following technical solution:

[0012] A multi-IC electromagnetic coupling decoupling guidance method based on system-level EMC data assets includes the following steps:

[0013] S1. Obtain user-specified IC combination information, which includes the identifiers of at least two ICs and their layout relationships in the target system; retrieve a baseline coupling coefficient matrix matching the IC combination information from the system-level EMC data asset; when the IC combination information is not found in the data asset, provide an estimated result based on similar IC combinations and generate a standardized requirement form to be sent to the data asset producer. The data asset producer generates corresponding IC-level and system-level data assets based on a microcircuit verification platform independent of the user's system board, realizing on-demand generation and delivery of data assets and continuously updating the system-level EMC data assets.

[0014] The coupling confidence is comprehensively evaluated based on factors such as IC model matching, frequency similarity, and historical project similarity; the comprehensive coupling contribution value is the sum of the coupling contribution values ​​of this IC to all other ICs in the combination; the coupling offset is the difference between the measured scenario coupling coefficient and the benchmark coupling coefficient.

[0015] S2. Based on the baseline coupling coefficient matrix, calculate the comprehensive coupling contribution value of ICs existing in the data asset to other ICs in the combination, and identify the dominant coupling path based on the decomposition coefficient of each path; for ICs not existing in the data asset, estimate based on the baseline coupling coefficient of similar ICs; record the calculation process of coupling contribution value, the basis for identifying dominant paths, the basis and reasons for threshold selection, and generate coupling analysis process data assets; the coupling analysis process data assets include timestamps, input conditions, key IC identifiers, dominant path types and coupling confidence levels, for traceability in the design process; key coupling paths are selected according to preset thresholds.

[0016] S3. Based on the key coupling paths and their dominant path types, generate decoupling instructions. These decoupling instructions include mandatory instructions and reserved instructions. Mandatory instructions are design rules recommended for execution based on high-confidence data or high-risk assessments. Reserved instructions are design measures reserved based on medium-to-low confidence data or uncertainty assessments. The decoupling instructions are output to the user's local design knowledge base in the form of structured data assets, including design parameter reference values, effect estimates, and applicable conditions. The effect estimates include the expected improvement reference value and fluctuation range. The reserved measures include optional filtering, shielding, or isolation positions pre-laid on the PCB, associated with applicable conditions.

[0017] S4. Store the decoupling instructions in the user's local design knowledge base as a callable data asset for subsequent IC layout design. The user's local design knowledge base is independent of the system-level EMC data assets and is used to store decoupling instructions, test data, and correction factors accumulated by the user based on their own projects, forming user-specific design experience assets.

[0018] Furthermore, the method is applied to EMC intelligent design instruments.

[0019] Furthermore, the method is implemented as a computer software product, including standalone applications or software plugins.

[0020] Furthermore, the IC identification information in step S1 includes one or more of IC model, package type, function type, and operating frequency; the matching includes exact match and similar match. When an exact match does not exist, similar ICs are recommended based on function type and operating frequency, and a coupling confidence level prompt is given.

[0021] Furthermore, the data requirement process in step S1 includes: automatically generating a standardized requirement sheet, which includes the model, package type, functional parameters, combination relationship and expected application scenario of each IC in the combination; sending the requirement sheet to the data asset producer; receiving the IC combination system-level data asset generated by the producer based on the requirement, updating it to the local data asset library and reassessing the coupling confidence.

[0022] Furthermore, step S2, identifying the dominant coupling path, includes: identifying the type of the dominant coupling path as power conduction, signal coupling, or space radiation based on the decomposition coefficients of each path, and recording this as the basis for path decision-making.

[0023] Furthermore, step S2, which involves screening key coupling paths, includes: calculating the comprehensive coupling contribution value of ICs in the data assets, screening out key ICs based on configurable thresholds, and marking the coupling paths between key ICs as key coupling paths.

[0024] Furthermore, the decoupling instructions generated in step S3 include: providing a priority processing direction based on the type of the dominant coupling path; when the dominant path is spatial radiation, the mandatory instruction is to prioritize spatial isolation or shielding, and provide a spacing reference range or shielding suggestion; when the dominant path is a power path, the mandatory instruction is to prioritize power filtering or decoupling, and provide a filter device selection reference or layout suggestion; when the dominant path is a signal path, the mandatory instruction is to prioritize optimizing the trace layout or grounding, and provide a trace rule reference.

[0025] Furthermore, the generation of decoupling instructions in step S3 also includes: calling the interference source location database to obtain the location and intensity information of the interference source, and combining it with the key coupling path to generate a targeted layout adjustment scheme.

[0026] Furthermore, when generating decoupling instructions in step S3, the measure sensitivity assessment module can be called to obtain the quantitative effects of different decoupling measures in order to optimize instruction generation.

[0027] Beneficial effects

[0028] First, this invention decomposes electromagnetic coupling between ICs into the contribution values ​​and proportions of power supply paths, signal paths, and spatial radiation paths through path decomposition of the reference coupling coefficient matrix. This allows engineers to understand not only the coupling strength but also the main propagation paths of the coupling. This path-level decomposition provides a clear decision-making basis for subsequent decoupling, helping to reduce the risks of over-design or under-design.

[0029] Second, this invention introduces coupling confidence to quantify the degree of matching between data assets and the current design scenario. When the confidence is high, decoupling instructions can be directly used as a design basis; when the confidence is low, decoupling instructions include more adequate safeguards to address potential deviations, thereby ensuring the reasonable use of the reference value of data assets.

[0030] Third, this invention transforms complex EMC analysis into quantifiable and executable design rules by generating decoupling instructions. The instructions are divided into mandatory and reserved instructions, ensuring both decoupling effectiveness and design flexibility. This allows ordinary hardware engineers to quickly accumulate EMC design experience in practice, gradually growing into engineers with professional capabilities, and effectively reducing the diversity of design results caused by differences in individual experience.

[0031] Fourth, this invention combines simulation data to generate a spatial radiation heat map, which visually displays the coupling hotspots and sensitive areas around the IC. Engineers can use this to make reasonable plans during the layout stage, such as identifying high-risk areas and reserving shielding positions, which helps reduce the risk of design modifications caused by spatial coupling in the later stages.

[0032] Fifth, this invention provides reference values ​​and fluctuation ranges for expected improvements by outputting effect predictions, enabling engineers to have a quantitative understanding of the risks of design decisions and facilitating the reasonable arrangement of verification and adjustment stages in project planning.

[0033] Sixth, this invention achieves adaptive optimization of data assets by collecting test data from actual scenarios, comparing it with expected results, and using this data to update local correction factors. As project experience accumulates, the local correction factors are continuously optimized, and the coupling confidence of subsequent similar scenarios gradually increases, forming a continuous evolutionary capability of experience iteration, thereby gradually improving the effectiveness of data asset usage and prediction accuracy.

[0034] Seventh, this invention stores decoupling instructions in the user's local design knowledge base, forming reusable experience assets. Subsequent designs can directly call upon successful instructions from similar scenarios, significantly improving design efficiency and quality.

[0035] Eighth, this invention can be integrated into EMC intelligent design instruments or implemented as computer software products (including standalone applications, software plug-ins, etc.), providing flexible application options for enterprises of different sizes and broadening the application scenarios of the technology. Attached Figure Description

[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0037] Figure 1 The method flowchart of the present invention.

[0038] Figure 2 This invention relates to the inter-module interaction relationships in EMC intelligent design instruments.

[0039] Figure 3 This invention is illustrated as an application diagram of a standalone software tool.

[0040] Figure 4 A schematic diagram of the application of decoupling instructions and the closed loop of empirical iteration.

[0041] Figure 5 Interaction diagram of the decoupling guidance module based on data assets. Detailed Implementation

[0042] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0043] The overall process of this invention includes: S1. Obtaining the IC combination information specified by the user, retrieving the benchmark coupling coefficient matrix matching the IC combination information from the system-level EMC data asset; if it does not exist, generating a standardized requirement form and sending it to the data asset producer, and providing an estimated result based on similar ICs; S2. If a matching benchmark coupling coefficient matrix exists, using the matrix directly for calculation; if it does not exist, using the estimated value based on similar ICs in step S1 for calculation, calculating the comprehensive coupling contribution value of each IC, identifying the dominant path based on path decomposition, recording the calculation process of the coupling contribution value, the basis for identifying the dominant path, the basis and reason for the threshold selection, generating coupling analysis process data assets, and determining the key coupling path; S3. Generating decoupling instructions based on the key path and dominant path types, the decoupling instructions include mandatory instructions, reserved instructions and effect estimates, and outputting them in the form of structured data assets; S4. Storing the decoupling instructions in the user's local design knowledge base for subsequent design calls.

[0044] The following three embodiments illustrate the specific applications of the present invention in different carriers and modes, and highlight how core concepts such as coupling confidence, decoupling instructions, effect prediction, and local correction factors are integrated into the actual decision-making process.

[0045]

Example 1

[0046] This embodiment uses the integration of the method of the present invention into an EMC intelligent design instrument as an example to illustrate its application in the design mode. It also demonstrates a high degree of coupling confidence and significant data asset reference value.

[0047] An engineer is designing a drone flight control system, which includes the following ICs:

[0048] IC1-1 (Main processor, model STM32F407, package LQFP144, operating frequency 168MHz).

[0049] IC1-2 (Power management IC, model TPS5430, package HTSSOP-14, operating frequency 500kHz).

[0050] IC1-3 (Wireless Communication IC, Model CC1101, QFN-20 Package, Operating Frequency 433MHz).

[0051] Engineers input the above IC combination information into the EMC intelligent design instrument, and the instrument retrieves the corresponding baseline coupling coefficient matrix and coupling confidence from the system-level EMC data assets.

[0052] Step S1: Obtain the baseline coupling coefficient matrix and coupling confidence level.

[0053] The reference coupling coefficient matrix acquired by the instrument contains path decomposition information. For example, the reference coupling coefficient of IC1-1 to IC1-3 is 0.8, with the power path contributing 0.4 (50%), spatial radiation contributing 0.3 (37.5%), and the signal path contributing 0.1 (12.5%). The coupling confidence level is 85%, indicating a high degree of matching between the data assets and the current design scenario.

[0054] Step S2: Quantify coupling contributions, identify dominant paths, and determine key coupling paths.

[0055] The instrument calculates the combined coupling contribution of each IC to other ICs based on the reference coupling coefficient matrix:

[0056] IC1-1 contributes 0.3 to IC1-2 and 0.8 to IC1-3, with a total value of 1.1;

[0057] IC1-2 contributes 0.3 to IC1-1 and 0.1 to IC1-3, with a combined value of 0.4.

[0058] IC1-3 contributes 0.8 to IC1-1 and 0.1 to IC1-2, for a combined value of 0.9.

[0059] A threshold of 0.6 was set to identify ICs whose overall contribution value exceeded the threshold as critical ICs, namely IC1-1 and IC1-3. The critical coupling path is the bidirectional path between IC1-1 and IC1-3. Further identification of the dominant path revealed that the coupling of IC1-1 to IC1-3 is primarily via the power path (50%), followed by spatial radiation (37.5%). The instrument recorded the calculation process of the coupling contribution value, the basis for identifying the dominant path, and the basis and rationale for the threshold selection, generating a coupling analysis process data asset.

[0060] Step S3: Generate decoupling instructions

[0061] Based on the dominant path, the instrument generates decoupling instructions, including mandatory instructions and reserved instructions:

[0062] Mandatory instructions (power path): Add a π-type filter at the output of IC1-1, using a 10μF and a 0.1μF capacitor connected in parallel, with the recommended ferrite bead model being BLM18PG121SN1; Add a local decoupling capacitor at the input of IC1-3, using a 4.7μF and a 0.01μF capacitor connected in parallel. Failure to execute these instructions results in a high risk level for the critical coupling path.

[0063] Reserved instruction (space radiation): Reserve a shielding pad above IC1-3, with a size of 8mm×8mm; it is recommended that the distance between IC1-1 and IC1-3 be controlled at more than 15mm. If subsequent testing finds that the 125MHz frequency point exceeds the limit, the shielding cover can be soldered.

[0064] Auxiliary suggestion: Place IC1-2 between IC1-1 and IC1-3 as isolation.

[0065] The instrument's simultaneous output effect is estimated as follows: "After implementing the mandatory command, the coupling contribution of IC1-1 to IC1-3 is expected to be reduced to below 0.4, with a fluctuation range of ±0.05; if the reserved command is further implemented, it is expected to be reduced to below 0.3."

[0066] Step S4: Storage and Application

[0067] Engineers adjust the layout in the PCB design software according to the instructions. The first version executes the mandatory instructions first, reserving space for shielding covers. Decoupling instructions are stored in the user's local design knowledge base.

[0068] Subsequent feedback and experience iteration: Product testing passed, with a measured coupling coefficient of 0.42 and a coupling offset of 0.38 from the baseline coupling coefficient. This data was used to update the local correction factor (fine-tuning the attenuation coefficient of the power path), thereby increasing the coupling confidence of IC1-1 and IC1-3 to 88% in subsequent similar projects.

[0069] This embodiment demonstrates that in high-confidence scenarios, the decoupling instructions are clear, the effect prediction is accurate, the reserved measures serve as insurance, and the experience iteration makes the data assets more and more accurate with use.

[0070]

Example 2

[0071] This embodiment uses the integration of the method of the present invention into an EMC intelligent design instrument as an example to illustrate its application in diagnostic mode. The coupling confidence is moderate, and the decoupling instructions contain more adequate reserve measures.

[0072] An engineer designed an industrial Ethernet switch that was found to have excessive radiation during testing. The system contains the following ICs:

[0073] IC2-1 (Switching chip, model BCM53115, package BGA-256, operating frequency 125MHz).

[0074] IC2-2 (PHY chip, model BCM5482, package QFN-56, operating frequency 25MHz);

[0075] IC2-3 (clock IC, model SI5338, package QFN-24, operating frequency 200MHz).

[0076] The engineer input the measured spectrum into the EMC intelligent design instrument. The instrument called the interference source location database and indicated that IC2-1 was the main interference source, contributing about 70% at the 125MHz frequency point.

[0077] Step S1: Obtain the baseline coupling coefficient matrix and coupling confidence level.

[0078] The instrument retrieves the baseline coupling coefficient matrix and path decomposition for three ICs from the system-level EMC data assets. For example, the baseline coupling coefficient of IC2-1 to IC2-3 is 0.75, with spatial radiation contributing 0.45 (60%), power path contributing 0.2 (26.7%), and signal path contributing 0.1 (13.3%). The coupling confidence level is 70%.

[0079] Step S2: Quantify coupling contributions, identify dominant paths, and determine key coupling paths.

[0080] Quantify the overall contribution value of each IC:

[0081] IC2-1 contributes 0.45 to IC2-2 and 0.75 to IC2-3, with a combined value of 1.2;

[0082] IC2-2 contributes 0.45 to IC2-1 and 0.2 to IC2-3, with a combined value of 0.65;

[0083] IC2-3 contributes 0.75 to IC2-1 and 0.2 to IC2-2, for a combined value of 0.95.

[0084] Based on the interference source localization results, the critical coupling paths are IC2-1 to IC2-3 and IC2-1 to IC2-2. A threshold of 0.4 was set, classifying both as critical paths. Dominant path analysis showed that the coupling between IC2-1 and IC2-3 is primarily spatial radiation (60%). The instrument recorded the calculation process of the coupling contribution value, the basis for identifying the dominant path, and the basis and rationale for threshold selection, generating a coupling analysis process data asset.

[0085] Step S3: Generate decoupling instructions

[0086] Based on the dominant path, the instrument generates decoupling commands. Due to the moderate coupling confidence, contingency measures are strengthened.

[0087] Mandatory instruction (spatial radiation): Move IC2-3 out of the near-field region of IC2-1, increasing the spacing from the current 8mm to over 20mm. Failure to execute this instruction results in a high risk level.

[0088] Reserved instructions: Reserve a shielding pad above IC2-3, with a size of 5mm×5mm; Reserve a grounding via isolation strip position between IC2-1 and IC2-3; Add a ferrite bead (model BLM18PG121SN1) to the power pin of IC2-1 as an optional measure.

[0089] Additional suggestion: Adjust the PCB stack-up structure and place the signal layers of IC2-1 and IC2-2 on different reference planes.

[0090] Instrument output performance forecast: "After implementing the mandatory space radiation command, the coupling of IC2-1 to IC2-3 is expected to decrease to below 0.45, with a fluctuation range of ±20%. If it still exceeds the standard after the first test, we can try to add a shielding cover and add grounding vias in turn, which is expected to further reduce it to below 0.3."

[0091] Step S4: Storage and Application

[0092] The engineers adjusted the settings as instructed: first, they increased the spacing and added magnetic beads, without installing a shielding cover. After testing, the radiation level dropped to near the limit but still exceeded it by 2 dB. Subsequently, the shielding cover was installed, and the radiation problem was effectively resolved. The intermediate data recorded during this diagnostic process was generated as a coupling analysis process data asset, including information such as coupling confidence levels.

[0093] Subsequent feedback and experience iteration: The measured scene coupling coefficient was 0.48, and the coupling offset was 0.27. This data was used to update the local correction factor (to correct the attenuation coefficient of the spatial radiation path), which improved the coupling confidence of IC2-1 and IC2-3 in subsequent similar switch projects to 78%.

[0094] This embodiment demonstrates that in a medium confidence scenario, decoupling instructions control risks by strengthening reserved measures, the effect prediction indicates uncertainty, and the deviation is gradually reduced through experience iteration.

[0095]

Example 3

[0096] This embodiment uses the method of the present invention as an independent software tool to illustrate its universality under different carriers, and the coupling confidence is low, with the decoupling instructions containing multi-level reserved measures.

[0097] An engineer is designing a 5G base station radio frequency unit. The system includes the following ICs:

[0098] IC3-1 (RF transceiver, operating frequency 3.5GHz);

[0099] IC3-2 (Power Amplifier, Operating Frequency 3.5GHz);

[0100] IC3-3 (Phase-locked loop, operating frequency 2.5GHz);

[0101] IC3-4 (Digital Processing IC, operating frequency 1.2GHz);

[0102] IC3-5 (Power Management IC, operating frequency 1MHz).

[0103] Engineers input the aforementioned IC combination information into standalone software, which then retrieves the corresponding baseline coupling coefficient matrix from locally stored system-level EMC data assets.

[0104] Step S1: Obtain the baseline coupling coefficient matrix and coupling confidence level.

[0105] The baseline coupling coefficient matrix obtained by the software contains path decomposition information. For example, the baseline coupling coefficient of IC3-1 to IC3-2 is 0.9, with spatial radiation contributing 0.6 (66.7%), signal path contributing 0.2 (22.2%), and power path contributing 0.1 (11.1%). The coupling confidence level is 60%. The software also generates a spatial radiation heatmap based on simulation data, showing the upper right corner of IC3-1 as a coupling hotspot and the lower left corner of IC3-2 as a sensitive area.

[0106] Step S2: Quantify coupling contributions, identify dominant paths, and determine key coupling paths.

[0107] The software quantifies the overall contribution value of each IC, and calculates it based on the corresponding values ​​in the benchmark coupling coefficient matrix:

[0108] IC3-1 contributes 0.9 to IC3-2, 0.7 to IC3-3, 0.3 to IC3-4, and 0.1 to IC3-5, for a total value of 2.0.

[0109] IC3-2 contributes 0.9 to IC3-1, 0.6 to IC3-3, 0.2 to IC3-4, and 0.1 to IC3-5, for a total value of 1.8.

[0110] IC3-3 contributes 0.7 to IC3-1, 0.6 to IC3-2, 0.4 to IC3-4, and 0.2 to IC3-5, for a total value of 1.9.

[0111] The overall score for IC3-4 is 1.2, and the overall score for IC3-5 is 0.7.

[0112] A threshold of 1.5 was set, with ICs 3-1, 3-2, and 3-3 identified as key coupling elements. Key coupling paths included those between IC3-1 and IC3-2, IC3-1 and IC3-3, and IC3-2 and IC3-3. Dominant path analysis showed that the coupling between IC3-1 and IC3-2 was primarily spatial radiation (66.7%). The software recorded the calculation process of the coupling contribution value, the basis for identifying the dominant paths, and the rationale for selecting the threshold, generating a coupling analysis process data asset.

[0113] Step S3: Generate decoupling instructions

[0114] Due to the low coupling confidence level and high decoupling risk index, the software generates decoupling instructions that include multiple levels of reserved measures:

[0115] Mandatory instruction (space radiation core): The IC3-1 and IC3-2 RF links are isolated by stripline and laid out at a 90-degree angle. Failure to execute this instruction will result in a high risk level.

[0116] First-level reserved instructions: Reserve a shielding pad above IC3-2, with a size of 10mm×10mm; Reserve grounding via array positions on both sides of the stripline.

[0117] Second-level reserved instructions: Move IC3-3 out of the hot zone and place it on the edge of the PCB, and add a grounding via array; at the same time, reserve the position of an optional metal isolation wall.

[0118] The third-level reserved instruction is to add LC filters for the three key ICs in the power distribution network, with an inductor of 47nH and a capacitor of 100pF, and reserve spare LC filter positions.

[0119] Level 4 reserved instructions: Adjust the stack-up structure, place the RF layer and digital layer on different reference planes; reserve optional absorbing material mounting area.

[0120] Software output effect prediction: "After implementing the mandatory instructions, the coupling between IC3-1 and IC3-2 is expected to decrease by about 40%, with a fluctuation range of ±30%. If the first test fails to meet the standard, the first to fourth level of reserved measures can be tried in sequence, and it is expected to gradually reduce to below 0.5.

[0121] Step S4: Storage and Application

[0122] Engineers implemented the necessary measures as instructed and reserved all alternative locations. Initial testing revealed that radiated emissions at the 3.5GHz frequency point exceeded the limit by approximately 5dB. After implementing the first-level reserved measures (installing a shielding cover), the test passed. The decoupling instructions were stored in the local design knowledge base.

[0123] Subsequent feedback and experience iteration: The measured coupling coefficient was 0.55, and the coupling offset was 0.35. This data was used to update the local correction factor (significantly correcting the attenuation coefficient of the spatial radiation path), increasing the coupling confidence of similar IC combinations in subsequent 5G RF unit projects from 60% to 72%.

[0124] This embodiment demonstrates that the implementation of the present invention as a standalone software tool can also effectively guide the design, verifying the versatility of the method.

[0125] Application Examples

[0126] 1. Application in EMC intelligent design instruments: Industrial controller layout optimization

[0127] In the design of a multi-IC circuit board for an industrial controller, engineers input IC combination information using an intelligent EMC design instrument. The instrument retrieves the corresponding baseline coupling coefficient matrix from the system-level EMC data assets, and after path decomposition and dominant path identification, generates decoupling instructions containing mandatory instructions and reserved measures. Based on this, engineers adjust the layout in the PCB design software, effectively reducing the number of design iterations.

[0128] 2. Application in standalone software tools: Diagnosis and optimization of communication equipment

[0129] A certain communication device emitted excessive radiation. Engineers input the measured spectrum into a separate software tool. The software then used an interference source location database to pinpoint the dominant interference source and generated targeted decoupling instructions based on the method of this invention. Engineers adjusted the layout and added safety precautions according to the instructions, and the retest passed. Throughout the process, the location results, decoupling instructions, and measured data were all recorded as traceable data assets, forming a closed-loop optimization.

[0130] 3. Iterative Design of Complex Systems: Accumulation of Knowledge Bases from Multiple Projects

[0131] In multiple 5G base station transceiver projects, the design team stores each decoupling instruction, implementation effect, and measured data in a local design knowledge base. As projects accumulate, the local correction factor is continuously optimized, and the coupling confidence level gradually improves. When a new project starts, historical instructions from similar scenarios can be quickly retrieved, significantly improving design efficiency.

[0132] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-IC electromagnetic coupling decoupling guidance method based on system-level EMC data assets, characterized in that, Includes the following steps: S1. Obtain the IC combination information specified by the user, the combination information including the identifiers of at least two ICs and their layout association in the target system; retrieve the baseline coupling coefficient matrix that matches the IC combination information from the system-level EMC data assets; S2. Based on the aforementioned benchmark coupling coefficient matrix, calculate the comprehensive coupling contribution of each IC to other ICs in the combination, and identify the dominant coupling path based on the decomposition coefficients of each path. Record the quantification process and intermediate results, and filter out key coupling paths based on preset thresholds; S3. Based on the key coupling paths and their dominant path types, generate decoupling instructions; the decoupling instructions include mandatory instructions and reserved instructions; the mandatory instructions are design rules recommended for execution based on high-confidence data or high-risk judgments, and the reserved instructions are design measures recommended for reservation based on medium- and low-confidence data or uncertainty assessments; the decoupling instructions are output to the user's local design knowledge base in the form of structured data assets, including design parameter reference values, effect estimates, and applicable conditions; S4. Store the decoupling instructions in the user's local design knowledge base as a callable data asset for subsequent IC layout design.

2. The method according to claim 1, characterized in that, The method is applied to EMC intelligent design instruments.

3. The method according to claim 1, characterized in that, The method is implemented as a standalone computer software product.

4. The method according to claim 1, characterized in that, The IC identification information in step S1 includes one or more of IC model, package type, function type, and operating frequency; the matching includes exact match and similar match. When no exact match exists, similar ICs are recommended based on function type and operating frequency, and a coupling confidence level prompt is given.

5. The method according to claim 1, characterized in that, The data requirement process in step S1 includes: automatically generating a standardized requirement sheet, which includes the model, package type, functional parameters, combination relationship and expected application scenario of each IC in the combination; sending the requirement sheet to the data asset producer; receiving the IC combination system-level data asset generated by the producer based on the requirement, updating it to the local data asset library and reassessing the coupling confidence.

6. The method according to claim 1, characterized in that, Step S2 involves identifying the dominant coupling path, which includes determining the type of dominant coupling path as power conduction, signal coupling, or space radiation based on the decomposition coefficients of each path, and recording this information as the basis for path decision-making.

7. The method according to claim 1, characterized in that, Step S2, which involves screening key coupling paths, includes: calculating the comprehensive coupling contribution value of ICs in the data assets, screening out key ICs based on configurable thresholds, and marking the coupling paths between key ICs as key coupling paths.

8. The method according to claim 1, characterized in that, The decoupling instructions generated in step S3 include: providing a priority processing direction based on the type of the dominant coupling path; when the dominant path is spatial radiation, the mandatory instruction is to prioritize spatial isolation or shielding, and provide a spacing reference range or shielding suggestion; when the dominant path is a power path, the mandatory instruction is to prioritize power filtering or decoupling, and provide a filter device selection reference or layout suggestion; when the dominant path is a signal path, the mandatory instruction is to prioritize optimizing the trace layout or grounding, and provide a trace rule reference.

9. The method according to claim 1, characterized in that, Step S3, which generates decoupling instructions, also includes: calling the interference source location database to obtain the location and intensity information of the interference source, and combining it with the key coupling path to generate a targeted layout adjustment scheme.

10. The method according to claim 1, characterized in that, Also includes: The decoupling instructions are applied to the actual IC layout design, and actual electromagnetic compatibility performance data is collected to obtain the scenario coupling coefficient. The coupling offset between the coefficient and the reference coupling coefficient is calculated and updated to the user's local design knowledge base for generating or updating local correction factors. Through multiple rounds of project accumulation, experience iteration is formed to improve the coupling confidence of subsequent similar scenarios.

Images