mog
New Member
Offline
Posts: 1
|
Dear Gentlemen I have some questions in the jitter analysis and phase noise analysis.
When I run a open-loop VCO in circuit level with "spectre transient noise" added, I can get one sequence of the VCO output clock periods which have random noise.
I use this sequence of periods to convert to the PSD of phases. By the period jitter definition, I let
% Compute Period Jitter Using PSD function phases_period=2*pi*periods/Tref; % Compute Power Spectral Density of Phase [Pperiod,f_period]=psd(phases_period,nfft,1/Tref,winLength,overlap,'linear');
And by the TIE jitter defintion, I use accucmulation as below:
% Compute TIE Jitter Using PSD function % The Cumulative Phase of Each Transition phases_tie=2*pi*cumsum(periods)/Tref; % Compute Power Spectral Density of Phase [Ptie,f_tie]=psd(phases_tie,nfft,1/Tref,winLength,overlap,'linear');
Finally the results show that the PSD of the TIE jitter would be the same with Spectre's Pnoise analysis result.
But by many articles, they always relate the phase noise with period jitter. I am really confused....which one is right?
|