Statistical Analysis of Coupled Circuit-EM Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods fail to accurately model the statistical behavior of high-frequency RF circuits, particularly in combining circuit and electromagnetic (EM) variables, due to limitations in existing analytical techniques that cannot handle non-Gaussian, correlated parameters representing EM passive components.
Innovation Solution
A methodology is developed to decouple circuit and EM components, characterizing them separately using Y-parameters and creating response surfaces for both, allowing for the computation of performance metrics as Gaussian random variables, facilitating rapid Response Surface Monte Carlo analysis to determine Probability Density Functions (PDFs) and yield measures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional analytical techniques are used to model circuit variables, then circuit performance can be analyzed, but electromagnetic (EM) variables and non-Gaussian correlated parameters cannot be accurately handled
Solution Approach 1:
The patent segments the coupled circuit-EM system into separate circuit variables and EM variables. Circuit variables are modeled using traditional statistical techniques while EM variables are handled through field solvers that generate response surfaces. This segmentation allows each type of variable to be treated with the most appropriate methodology, resolving the contradiction between reliability for circuit analysis and adaptability for EM analysis.
Solution Approach 2:
The patent introduces response surfaces as an intermediary between EM field solvers and statistical analysis. The field solvers generate response surfaces that approximate EM behavior, which then serve as input to the statistical PDF calculation. This intermediary enables the integration of EM variables into the statistical framework without requiring direct modification of either the field solver or statistical tools.
2Measurement precision
If field solvers are used to characterize distributed interconnect structures, then EM effects are captured accurately, but integration with SPICE-like circuit simulators becomes complex
Solution Approach 1:
The patent extracts EM components (distributed interconnect structures, passive components) from the full circuit-EM system and characterizes them separately using field solvers. The EM behavior is captured through response surfaces that describe the relationship between geometric parameters and EM performance. These extracted EM models are then integrated with circuit simulators through a unified statistical framework, reducing the complexity of direct integration while maintaining accuracy.
Solution Approach 2:
The patent changes the parameter representation by using response surfaces that relate geometric parameters to EM performance metrics. Instead of directly integrating field solvers with circuit simulators, the methodology transforms EM behavior into parameter-based models that can be more easily integrated. This parameter transformation simplifies the integration process while preserving the accuracy of EM effects modeling.
3Manufacturing precision
If statistical analysis is performed after design enters manufacturing stage, then empirical results from production testing are obtained, but development time and costs increase substantially
Solution Approach 1:
The patent performs statistical analysis of circuit-EM systems during the design phase rather than after manufacturing begins. By calculating PDFs of performance metrics before the design is finalized, designers can identify yield issues and modify designs proactively. This preliminary statistical evaluation prevents costly post-manufacturing redesigns and accelerates time-to-market.
Data Source
AI summary
A method and system for combining the process variations in circuits and distributed interconnect-based electromagnetic (EM) objects in order to capture a statistical behavior of overall circuit performance parameters. In an exemplary approach, a coupled circuit-EM system is decoupled at the points where the EM objects connect to the circuit portion, and circuit ports are defined at those points. The sources of variation are identified and used to determine Y-parameters for the ports with EM elements and for all EM elements based on the SPICE-like and EM full-wave simulations. A response surface is generated for each variable and is used to extract circuit and EM parameters by generating many random vectors representing combinations of the random variables. These Y-parameters are merged to produce a probability density function (PDF) of one or more performance metrics for the electronic device or system.