Hello people

I need to find the phase noise contributed by a phase/frequency detector + charge pump combination which are part of a frequency synthesizer. When the synthesizer is locked, both the inputs to the PFD are of the same frequency and hence the beat frequency is zero (albeit there will be reference feedthrough). But I cannot do a PSS/Pnoise analysis with a 0 beat frequency. How do I proceed?

I tried measuring phase noise while having a finite frequency difference too, but the simulation never seems to converge however much I change the tstab value or if I make it conservative or liberal. BTW I use cadence spectre's analog environment for the simulations.

Thanks in advance

But Frank, since charge pump is shaped by the low pass filter in the loop, wont I be interested in noise near DC rather than at the reference frequency (since reference frequency is many times higher than loop bandwidth, and any noise at these frequencies will be suppressed anyway)?

OK. I'll try that. Thank you Frank.