Phase Noise Measurement Using Spur-Selective Periodic Jitter Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Phase noise test instruments struggle to accurately determine peak phase jitter due to spurs, as they either incorrectly compute root mean square values or rely on inaccurate model-fits, leading to significant errors in peak-peak periodic jitter estimation.
Innovation Solution
A signal analysis system that transforms and filters digitized radio frequency signals, detects spurs, sets frequency bins to zero, and performs inverse transformations to obtain accurate peak periodic jitter by demodulating phase signals and computing average slopes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If RMS value of all spurs is taken to compute RMS phase jitter, then computation is simplified, but measurement precision deteriorates due to incorrect periodic jitter estimation
Solution Approach 1:
The patent segments the phase noise spectrum into individual spur components and processes each spur separately by detecting its frequency, amplitude, and phase. This allows accurate reconstruction of the time-domain periodic jitter waveform from individual spur contributions, rather than computing a single RMS value that loses phase information.
Solution Approach 2:
The patent introduces an intermediary process that transforms frequency-domain spur measurements into time-domain periodic jitter waveforms through inverse Fourier transform. This intermediary transformation enables accurate peak-to-peak periodic jitter measurement by preserving phase relationships, bridging the gap between frequency-domain measurements and time-domain jitter characterization.
2Adaptability or versatility
If model-fit in time domain is used to separate random jitter from periodic jitter, then separation is achieved, but measurement precision deteriorates due to approximation errors
Solution Approach 1:
The patent replaces the mechanical model-fitting approach with a direct mathematical transformation method. Instead of using Dual-Dirac model fitting based on noise statistics, the patent uses inverse Fourier transform to directly compute the periodic jitter waveform from measured spur components, eliminating approximation errors inherent in model-based separation.
Solution Approach 2:
The patent changes the measurement parameters from statistical noise characteristics to direct spur component parameters (frequency, amplitude, phase). By measuring and transforming individual spur parameters rather than fitting statistical models to noise, the patent achieves accurate periodic jitter determination without relying on approximations.
3Device complexity
If AM rejection is limited to 18-25 dB, then device complexity is reduced, but measurement precision deteriorates when significant AM spurs are present
Solution Approach 1:
The patent inverts the traditional approach by not attempting to suppress AM spurs through rejection, but instead by measuring both AM and PM spurs separately and processing only the PM components for periodic jitter calculation. This inversion eliminates the need for high AM rejection while maintaining measurement accuracy.
Solution Approach 2:
The patent extracts only the phase modulation spur components from the combined AM/PM signal spectrum, separating them from amplitude modulation spurs. By extracting and processing only the relevant PM spur information for periodic jitter calculation, the patent avoids the impact of AM spurs without requiring complex rejection mechanisms.
Data Source
AI summary
A signal analysis system is disclosed. The signal analysis system to: transforms a digitized radio frequency signal into a first transformed digital signal; digitally downconverts the digitized radio frequency signal to a first baseband signal; filter the first baseband signal into a first filtered signal; transforms the first filtered signal into a second transformed digital signal; computes a first phase noise spectrum from the second transformed digital signal, detects spurs and frequencies corresponding to the spurs in the first phase noise spectrum, and set frequency bins not including spurs to zero to generate a modified second transformed digital signal; inversely transforms the modified second transformed digital signal into a first complex time-domain baseband signal; demodulates a first phase signal in the first complex time-domain baseband signal to obtain a first demodulated signal; and obtains an average of slopes and frequency error from the slopes of the first demodulated signal.