Digital PLL Linear Prediction for Spur and Jitter Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital phase-locked loops (DPLLs) suffer from spurious energy introduction due to time-to-digital converter gain error, non-linearity, and phase error, leading to undesirable periodic tones and increased jitter, which can cause mixing of unwanted signals and emission mask violations.
Innovation Solution
A linear prediction technique is employed to determine the location of spurious content in DPLLs using an all-zero filter with P weights corresponding to P zeros at frequencies of P spurs, suppressing these spurs without increasing the noise floor, and adapting to different loop configurations and environmental changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional digital phase-locked loop components (time-to-digital converter, fractional divider, phase accumulator) are used, then clock signal generation is achieved, but spurious energy and periodic tones are introduced
Solution Approach 1:
The patent applies linear prediction filtering to convert the harmful spurious energy into beneficial signal cancellation. The filter predicts and subtracts the periodic tones from the clock signal, transforming the harmful interference into a cleaned clock output. This resolves the contradiction by maintaining reliable clock generation while eliminating spurious energy through active cancellation rather than passive filtering.
Solution Approach 2:
The linear prediction filter acts as an intermediary component between the time-to-digital converter and the clock output. It processes the intermediate signal to remove spurious content before final output, serving as a mediator that preserves the functionality of existing DPLL components while eliminating their harmful effects. This allows the system to maintain clock generation reliability without directly modifying the core DPLL architecture.
2Object-generated harmful factors
If filtering techniques are applied to suppress spurs, then spurious tones are reduced, but system complexity increases
Solution Approach 1:
The patent uses parameter-based linear prediction filtering where the filter coefficients are determined by analyzing the spectral characteristics of the spurious tones. By changing the filter parameters (coefficients) based on the specific spur frequencies present, the system achieves effective suppression without requiring complex adaptive structures. This resolves the contradiction by using simple filter architecture with adjustable parameters rather than complex adaptive filtering algorithms.
3Object-generated harmful factors
If traditional spur suppression methods are used, then some spurs are reduced, but noise floor increases
Solution Approach 1:
The linear prediction filter selectively cancels only the spurious tones while preserving the legitimate clock signal and noise floor. By predicting and subtracting only the periodic components identified through spectral analysis, the method reduces spur amplitude without amplifying or adding noise. This resolves the contradiction by achieving spur suppression through precise spectral targeting rather than broad-spectrum filtering that would affect the noise floor.
Data Source
AI summary
A technique uses linear prediction to determine the location of spurious content in a digital phase-locked loop and suppresses the spurious content from propagating to the clock output. In at least one embodiment, the technique implements an iterative (e.g., recursive) computation.