The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> question about extraction phase noise parameters of pfd-cp and divider
https://designers-guide.org/forum/YaBB.pl?num=1217324514

Message started by noiseless on Jul 29^th, 2008, 2:41am

Title: question about extraction phase noise parameters of pfd-cp and divider
Post by noiseless on Jul 29^th, 2008, 2:41am

As expressed in Ken's “Predicting the phase noise and jitter of PLL-based frequency synthesizers”, in order to simulate in the phase domain, pfd-cp and divider phase noise parameters should be extracted first, then use it to do the simulation. However, I am confused about how to do it will get right result.
1.for example, for the divider, we know use jitter option of the PNoise to do the jitter simulation. But how to do to get phase noise? I tried to use equation (35) to do the compare, which tdomain output noise as Sn(f), Jee result as SΦ(f). However, they are not equal.
2. for the pfd-cp simulation, I use PNoise to do the simulation, the script is as below
analysis('pnoise ?sweeptype "relative" ?relharmnum "1" ?start "100"
?stop "50M" ?dec "30" ?maxsideband "30" ?oprobe "/IPRB0"
?iprobe "" ?refsideband "" )
analysis('pss ?fund "10M" ?harms "10" ?errpreset "moderate"
?tstab "200n" )
I use Iprobe to sense the current noise, which connected to a voltage source to bias the DC voltage. However, the result is somewhat weird (see the attached pic). There is some spurs in the black circle and there is no apparent noise bandwidth there as shown in Ken's paper. My setup is right or not? appreciate any comments

Title: Re: question about extraction phase noise parameters of pfd-cp and divider
Post by Ken Kundert on Jul 29^th, 2008, 1:36pm

2. Those peaks are associated with the harmonics of the oscillation. The phase noise is mixed up not just to the fundamental, but also to every harmonic. You sweep range overlaps several harmonics.

-Ken

Title: Re: question about extraction phase noise parameters of pfd-cp and divider
Post by noiseless on Jul 29^th, 2008, 6:51pm

Ken, appreciate your reply
for 2, do you have idea why there is no corner for the noise bandwidth?

for 1, I found that what get from Jee is not SΦ(f), which unit is sec/sqrt(Hz). So in order to get SΦ(f), SΦ(f) should be equal to (2πfo)^2*Jee. Am I right?

Title: Re: question about extraction phase noise parameters of pfd-cp and divider
Post by noiseless on Aug 3^rd, 2008, 10:17pm

In the Page 45 of Ken's "pnoise+jitter" papaer about matlab script for computing SΦ(Δf), cumfum function is used to compute phase. Here I am confused. periods variable is for jitter, which is random here. Fourier transform of autocorrelation of random is PSD. why do not use autocorrelation function but cumsum function?

actually, I tried autocorration function, the result seems not right. However I do not know the reason. Can anyone explain it? appreciate