Stats

  • Locked Locked
  • Replies 1
  • Subscribers 126
  • Views 13700
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Transient Analyses

analogy

Defining noise during transient simulation I found this (spectre -h tran):

52      noisefmax=0 Hz    The bandwidth of pseudorandom noise sources. A valid (nonzero) noisefmax turns on the noise sources during transient analysis. The maximum time step of the transient analysis is limited to 1/noisefmax.
53      noisescale=1      Noise scale factor applied to all generated noise. Can be used to artificially inflate the small noise to make it visible above transient analysis numerical noise floor, but it should be small enough to
                          maintain the nonlinear operation of the circuit .
54      noiseseed         Seed for the random number generator. Should be positive integer. Specifying the same seed allows you to reproduce a previous experiment.
55      noisefmin (Hz)    If specified, the power spectral density of the noise sources will depend on frequency in the interval from noisefmin to noisefmax. Below noisefmin the noise power density is constant. The default value is
                          noisefmax, so that only white noise is included by default, and noise sources are evaluated only at noisefmax for all models. 1/noisefmin cannot exceed the requested time duration of transient analysis.
56      noisetmin (s)     Time interval between noise source updates. Default is 1/noisefmax. Smaller values will produce smoother noise signals, but will reduce time integration step.
57      noiseupdate=fmax  Forces evaluation of bias-dependent device noise sources at each time step, even if it is smaller than noisetmin value. .Possible values are fmax or step.

We know that active device noise is defined by BSIM parameters tnoimod (thermal) and fnoimod (flicker) from technology model files.However it seems that choice of noisefmax and noisefmin defines bandwidth while noiseseed affects randomness.But how does transient noise analyses interact with spectre to shape device noise ?


Hope this doesn't sound confusing.

Thanks

  • Let me give a simplified high level overview of this:

    • If you specify noisefmax only, then at each time point it will find the magnitude of the noise at that frequency in each noise source in the device (which are described in the frequency domain). Then it will generate random voltages at each noise source which give the same PSD (which is white in this case).
    • If you also specify noisefmin, then at each time point it will calculate the noise frequency response between noisefmin and noisefmax (based on that bias point) and then generate random numbers weighted such that they fit this PSD (sort of)
    • noiseseed just adjusts the starting point for the random number generator. It's primarily there so you can repeat experiments or perturb the random number generator to get different results (although equally random).

    Overall, it's using the noise equations in the device models to figure out the shape of the noise at each noise source, and that information is then used to generate random numbers in the time domain which give the same PSD within the limits you've set.

    Does that help?

    Regards,

    Andrew.

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.