Noise Parameter Measurement Using Device Model Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise parameter measurement methods are limited by significant scatter in data when varying measurement parameters like frequency or bias, leading to inaccurate representation of device noise performance due to sensitivity and error in measurement processes.
Innovation Solution
Utilizing device models that incorporate known information about device performance, such as smooth behavior with frequency, to directly determine noise parameters from raw noise data, reducing the number of variables to be measured and incorporating over-determined data for error reduction, rather than relying solely on independent measurements at each parameter value.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If independent measurements are performed at each measurement parameter value (frequency, bias, temperature), then complete noise parameter data is obtained for each condition, but significant scatter appears in the data due to measurement sensitivity and error
Solution Approach 1:
The patent performs preliminary smoothing of noise parameter data as a function of measurement parameters (frequency, bias, temperature) before using the smoothed data to extract device model parameters. This preliminary action removes scatter and measurement errors in advance, ensuring that the subsequent parameter extraction is based on reliable, consistent data rather than raw measurements with significant scatter.
Solution Approach 2:
The patent introduces smoothed noise parameter data as an intermediary between raw measurements and device model parameters. The smoothing process acts as a mediator that transforms scattered measurement data into reliable input for model extraction, eliminating the need to directly use noisy raw data while preserving the essential device characteristics.
2Productivity
If noise parameters are extracted directly from raw noise data without smoothing, then measurement data is used directly, but the scatter in data leads to inaccurate representation of device noise performance
Solution Approach 1:
The patent applies preliminary smoothing to noise parameter data before extraction of device model parameters. This preprocessing step eliminates scatter and measurement errors while preserving the underlying device characteristics, ensuring accurate parameter extraction without requiring repeated measurements.
Solution Approach 2:
The patent replaces direct measurement-based parameter extraction with a model-based extraction approach using smoothed data. Instead of relying on raw measurement data with inherent scatter, the system uses a device model fitted to smoothed noise parameters, substituting the direct measurement path with a more reliable modeling path.
3Reliability
If smoothing is applied to noise parameter data, then scatter is reduced and data consistency is improved, but additional processing time and complexity are introduced
Solution Approach 1:
The patent applies smoothing as a preliminary processing step to noise parameter data before model parameter extraction. This approach improves data consistency and reliability by removing scatter, and the smoothed data serves as robust input for subsequent parameter extraction, justifying the additional processing step.
4Measurement precision
If smoothing is applied to noise parameter data, then scatter is reduced and accurate representation of device performance is achieved, but additional processing steps are required
Solution Approach 1:
The patent performs preliminary smoothing of noise parameter data to eliminate scatter and measurement errors before extracting device model parameters. This preprocessing ensures that the parameter extraction is based on accurate, reliable data, improving measurement precision despite the additional processing step.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods of measuring and determining noise parameters. An exemplary method measures noise data and determines element values of a device noise model for a device under test (DUT), using a test system including an impedance tuner coupled to an input of the DUT for presenting a controllable variable impedance to the DUT and a noise receiver coupled to an output of the DUT. Noise data is measured as a function of at least one measurement parameter. The measured data includes raw noise data read from the noise receiver, and is used to determine element values of the device noise model. The system may include a database of device models