Stats

  • Replies 6
  • Subscribers 131
  • Views 851
  • Members are here 0

Fundamental question on edge phase noise from sampled PNOISE

Yuto Lau

Hi,

I have a few fundamental questions on edge phase noise and sampled PNOISE sim. Could you please give me some idea or point me to any helpful material.

My question is a below, 

Phase noise in dBc/Hz with time average PNOISE, the sweep frequency range is usually set as "relative to fundamental tone". And it makes sense that phase noise is referred to fundamental tone power.

In sampled PNOISE sim for driven circuit, we set "absolute sweep frequency"  up to Nyquist of the clock. Because output noise is sampled by clock period in the measurement (This part also makes sense to me). But edge phase noise is plot in unit "dBc/Hz". What is the carrier power in this measurement? Is it still the fundamental tone of the clock?

Thanks and regards,

Yutao

  • Frank Wiedmann

    Edge phase noise takes the jitter from the sampled pnoise analysis and converts it to the equivalent phase noise in dBc/Hz of a sinusoidal oscillator. Design equations for PLLs etc. often use this phase noise value as an input. A sinusoidal oscillator with a phase noise equal to the edge phase noise value would exhibit the same jitter as given by the result of the sampled pnoise analysis.

    You might also want to take a look at Tawna's posts at community.cadence.com/.../how-to-derive-edge-phase-noise-from-output-noise-in-sampled-pnoise-simulation

  • Yuto Lau
    in reply to Frank Wiedmann

    Hi Frank,

    Thanks for your explanation.

    Based on your comment, as edge phase noise is the phase noise of equivalent sinusoidal oscillator exhibiting the same jitter,

    How should we understand the meaning of "frequency" in x-axis?

    Should it be the offset frequency "relative" to "clock frequency" as well as the "sampling frequency"? However, it is a absolute frequency sweep in PNOISE.

    Thanks and regards,

    Yutao 

  • Frank Wiedmann
    in reply to Yuto Lau

    For edge phase noise, the frequency axis is the offset relative to the oscillator frequency. The (SSB) phase noise according to the traditional definition is not the spectral density of the phase variation but the value that you get on a spectrum analyzer (or in a pnoise average analysis) at a certain offset from the oscillator frequency.

  • Yuto Lau
    in reply to Frank Wiedmann

    Hi Frank,

    Appreciate your reply.

    1. Why do you say "(SSB) phase noise is NOT the spectral density of phase variation in traditional definition"? Is it because it is the ratio between noise spectra density over carrier power? Or any other reason?

    2. Based on your reply, "absolute frequency sweep range" in sampled PNOISE sim is to determined how much freq. range noise we would like to include in the periodic sampling. While in edge phase noise plot, It is a presentation by a sinusoidal oscillator model achieving equivalent amount jitter. Those two frequency range is NOT the same thing?

    Thanks and regards,

    Yutao

  • Frank Wiedmann
    in reply to Yuto Lau
    Yuto Lau said:
    1. Why do you say "(SSB) phase noise is NOT the spectral density of phase variation in traditional definition"? Is it because it is the ratio between noise spectra density over carrier power? Or any other reason?

    It was defined like this because comparing the value at some frequency offset to the maximum of the peak is simply the easiest way to read a value on a spectrum analyzer.

    Yuto Lau said:
    2. Based on your reply, "absolute frequency sweep range" in sampled PNOISE sim is to determined how much freq. range noise we would like to include in the periodic sampling. While in edge phase noise plot, It is a presentation by a sinusoidal oscillator model achieving equivalent amount jitter. Those two frequency range is NOT the same thing?

    The frequency sweep in a sampled pnoise analysis should always be from (almost) zero to half of the pss fundamental (assuming the default sampleratio=1). How much noise you include depends on the maxsideband parameter of the pnoise analysis (or on maxacfreq if you use pnoisemethod=fullspectrum). The equivalent edge phase noise of the sinusoidal oscillator gives you the value that you would read on a spectrum analyzer at a certain offset from the oscillator frequency (as I already mentioned).

  • Yuto Lau
    in reply to Frank Wiedmann

    Hi Frank,

    Really appreciate your patient and detailed explanation.

    Now I have a better understanding on edge phase noise from sampled PNOISE sim.

    Regards,

    Yutao

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.