Evaluation of PLL noise by transient

Dear All:

I realize that it is difficult, or impossible, to get a explicit result of PLL phase noise through Spectre transient simulation, but we all note that the rms-jitter is the integration of PLL phase noise from DC to infinity. Most probably the way we did to evaluate the PN of PLL is through the math modeling in MATLAB , in which each noise contribution and shaping can be explicitly observed. However, if we want to try something new while the PLL noise model is not clearly known now, this method could be difficult. Fortunately, similar to clock and data recovery circuit (CDR), we always like to observe the eye diagram in transient to evaluate its jitters performance. As such, can we employ the similar way to observe the eye diagram of a synchronized clock when the PLL is locked so that the phase noise can be intuitively compared?

Thanks in advanced!

NovelPanda said:

As such, can we employ the similar way to observe the eye diagram of a synchronized clock when the PLL is locked so that the phase noise can be intuitively compared?

Click to expand...

Almost, impossible.

First, there is no device noise in conventional transient analysis.
Second, even if you use transient noise analysis, it requires very long stop time to evaluate phase noise of low offset frequency.

Even if your concern is rms jitter not PSD, you have to take very long stop time to include phase noise of low offset frequency effects.