Smith Chart Spiral Extrapolation for Missing Low-Frequency Network Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current measurement equipment typically cannot provide network parameter data at frequencies below 45 MHz, leading to inaccuracies in characterizing low-frequency behavior of passive electrical networks, which is critical for signal integrity analysis in digital PCBs and packaging, as existing extrapolation techniques often rely on assumptions that may not be accurate.
Innovation Solution
The use of a Smith Chart technique to extrapolate frequency domain network performance information by analyzing measured high-frequency data, employing spiral pattern extrapolation to predict lower frequency data points, ensuring accurate characterization down to DC frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurement equipment is used to obtain network parameter data, then measurement precision is improved, but the frequency range is limited to above 45 MHz, causing loss of information at lower frequencies
Solution Approach 1:
The patent performs preliminary action by measuring network parameters at high frequencies first, then uses spiral extrapolation to predict low-frequency behavior before actual low-frequency measurements would be needed. This allows the system to prepare complete frequency domain data (including DC to high frequency) in advance for accurate inverse Fourier transformation to time domain waveforms.
Solution Approach 2:
The patent introduces an intermediary mathematical model (spiral extrapolation based on Smith Chart theory) that bridges the gap between measurable high-frequency data and unmeasurable low-frequency data. This intermediary allows information to be transferred from the high-frequency measurement domain to the low-frequency prediction domain, filling the data gap without requiring direct low-frequency measurements.
2Adaptability or versatility
If extrapolation techniques are used to predict low frequency data, then coverage of frequency range is improved, but accuracy deteriorates due to reliance on assumptions about data points near DC
Solution Approach 1:
The patent changes the parameter basis for extrapolation from frequency-domain assumptions (which fail near DC) to impedance-domain geometry (Smith Chart spiral patterns). By transforming the problem from predicting frequency responses to following impedance trajectory patterns, the extrapolation maintains accuracy across the entire frequency range including DC, as impedance behavior follows more predictable geometric patterns than raw frequency responses.
3Productivity
If standard Fourier transform is used for time domain conversion, then signal processing capability is improved, but accuracy deteriorates due to missing DC and low frequency information
Solution Approach 1:
The patent performs preliminary action by pre-completing the frequency domain signature with extrapolated DC and low-frequency data before the inverse Fourier transformation is executed. This ensures that when the standard IFFT algorithm runs, it receives complete frequency information from DC to high frequency, eliminating the need for special handling or approximation during the transform itself while maintaining waveform accuracy.
Data Source
AI summary
A method is provided to use a Smith Chart technique to obtain frequency domain network performance information corresponding to a passive network including one or more passive devices comprising: receiving first data representing a first Smith Chart plot of coefficients representing measured mismatch between a source impedance of a network and a load impedance of the network for higher frequency components; and extrapolating a predicted substantially spiral shaped second Smith Chart plot of coefficients based upon the first data, which includes a coefficient representing predicted mismatch between the source impedance of the network and the load impedance of the network for lower frequency components.