The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
https://designers-guide.org/forum/YaBB.pl?num=1251320771

Message started by BillH on Aug 26^th, 2009, 2:06pm

Title: Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
Post by BillH on Aug 26^th, 2009, 2:06pm

Hello,

This question is related to the Jitter simulation from spectre pnoise. Is the spectral density of the Jitter (Jee) a single-sideband (SSB) density or a double-sideband(DSB) result? Another way of looking at the question is, if I want to get the total integrated noise from the spectral density would integrate the spectral density of the jitter and take that number or do I need to multiply by 2?

For a phase noise simulation, assuming that one is looking at a relative harmonic the result for phase noise is SSB - meaning that to get the integrated phase noise one needs to integrate the phase noise density then multiply by 2 to account for the other sideband. But I'm not sure if holds true for jitter.

To be more specific on the simulation, I have run a pss simulation then a pnoise simulation with noise type: jitter. Command was:

analysis('pnoise ?sweeptype "relative" ?relharmnum "1" ?start "1000" ?stop "10M" ?dec "1" ?maxsideband "50" ?p "/vop_ndiv" ?n "/gnd!" ?oprobe "" ?iprobe "" ?refsideband ""
?noisetypeUI "jitter" ?crossingdirection "rise" ?thresholdvalue "0.5" ?solver "turbo" )

Then to plot the spectral density of jitter, I did the following from the Analog Design Environment (ADE):

Results
->Direct Plot
->Main Form
->Analysis pnoise jitter
->Function Jee

This results in calculator expression:
_drplJitter(?result "pnoise_pmjitter" ?unit "Second" ?k 1 ?event 0))
So the plot shows the spectral density rms jitter per Hz. But then, is this density single-sideband or both sidebands?

Thanks for any advice.

Title: Re: Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
Post by Frank Wiedmann on Aug 27^th, 2009, 1:28am

I don't think that jitter/Hz is a very useful result. One is usually interested in jitter after a certain time (or a certain number of cycles) or in long-term jitter (for time going towards infinity). To get the long-term rms jitter, you have to integrate without multiplying by 2. You might want to have a look at http://www.designers-guide.org/Forum/YaBB.pl?num=1224609785 for some further information.

Title: Re: Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
Post by BillH on Aug 27^th, 2009, 10:58am

Thanks Frank.

I'm actually interested in simulating the phase noise of a frequency divider in a PLL. Ultimately, I'm interested in a phase noise number (in terms of dBc/Hz) which represents the jitter in the edge transition at the divider output (similar to what's described in http://www.designers-guide.org/Analysis/divnoise.pdf, or in chap. 9 of http://www.designers-guide.org/Analysis/PLLnoise+jitter.pdf) except I believe at the time of that article the "jitter" analysis method wasn't available). As you've pointed out in other posts the "jitter" method is a special case of strobed analysis where you're looking for a threshold crossing.

So ultimately I'm trying to get the equivalent phase noise of the frequency divider in dBc/Hz. Jitter can be converted to phase noise by multiplying by 2pi*freq but then there is still the question about single-sideband vs. double-sideband, i.e. do I need to subtract 3dB?

Title: Re: Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
Post by Frank Wiedmann on Aug 27^th, 2009, 3:23pm

That probably depends on what you mean by "phase noise". Take a look at section 3.2 and especially equation 27 of http://www.designers-guide.org/Analysis/PLLnoise+jitter.pdf.

Title: Re: Is the spectral density of jitter (Jee) in Cadence SSB or DSB?
Post by BillH on Aug 28^th, 2009, 10:29am

Thanks Frank.

What I'm referring to as "phase noise" would be the term on the left side of the equation 27 you refer to.

L(deltaF)

I would call this the "single-sideband" phase noise (I'm not 100% certain that terminology is universally used so please feel free to correct my terminology if necessary). I believe in practice that terminology and meaning for phase noise tends to be used because it's what you see on the spectrum analyzer in the lab.

Thanks again.