The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Measurements >> Phase Noise and Jitter Measurements >> Insufficient number of harmonics in pss affecting pnoise jitter
https://designers-guide.org/forum/YaBB.pl?num=1474028746

Message started by manikanta on Sep 16^th, 2016, 5:25am

Title: Insufficient number of harmonics in pss affecting pnoise jitter
Post by manikanta on Sep 16^th, 2016, 5:25am

Hi,

I am doing a simple experiment on pss and pnoise analysis.
My DUT is simple RC Low pass filter ckt where R=1 Ohm and C=1nF and BW = 159M Hz.
Input frequency of the ckt = 10M Hz(pss frequency).
I am measuring the jitter noise at the output across cap and it should be kT/C = 4.14pV
I have varied the number of harmonics from 20 to 1000 and no of sidebands = 2000.
Maxacfreq is set to default. Output noise numbers are in the image.

What is actually happening when there are insufficient number of harmonics in the pss simulation?
Why do the noise numbers go high when number of harmonics in pss are very low.

Note: I am aware that I need to set maxacfreq to no_of_sidebands*no_of_harmonics to get correct noise numbers. But trying to understand the working of pnoise on pss.

Title: Re: Insufficient number of harmonics in pss affecting pnoise jitter
Post by Ken Kundert on Sep 16^th, 2016, 2:24pm

You are chasing a side effect.
You only need to specify harmonics if you want to look at the PSS results in the frequency domain.
Spectre does not require you to set the number of harmonics, and generally you would only do it if you wanted to look at the spectra. However, for reasons I do not understand, ADE requires a value.

When you increase the number of harmonics, Spectre will shrink maxstep accordingly. Just as it does if you set maxacfreq. When you up the number of harmonics you see accuracy increasing because it is in effect increasing maxacfreq for you, but at the cost of computing more harmonics in the spectra of all the node voltages.

Just set maxacfreq.

-Ken

Title: Re: Insufficient number of harmonics in pss affecting pnoise jitter
Post by sheldon on Oct 28^th, 2016, 11:55pm

It also seems like you are using maxsidebands. If you use fullspectrum pnoise, this should not be required