Stats

  • Replies 1
  • Subscribers 126
  • Views 1433
  • Members are here 0

Reducing Numerical Noise for S-Parameter Simulation

sgcad

Hi

While simulating the phase variation of a variable attenuator for a phased array system, I've observed unexpectedly high numerical noise as shown below:

The phase variation (or already the direct plot of the phase) shows random jumps of about 2 degrees whose position depend on the sweep step size, to some extent on the global reltol setting, and whether I use APS or native Spectre. This simulation is done with a regular small-signal S-parameter analysis. The circuit is very simple, the operating points seem to be correct (nothing in a strange region), and the PDK is for a mature technology from a large manufacturer.

Besides the global retol setting, is there a Spectre parameter that can affect the numerical accuracy of a S-parameter simulation and that I could tighten?

Thanks very much for your reply!

  • Without more details, this is rather difficult to debug. There might be slight differences between the dc operating points. You could place an explicit dc analysis before the sp analysis; this will give you more options to control the accuracy of the dc result (which then should automatically be reused for the sp analysis). The option rebuild_matrix=yes has also been useful for us in the past to avoid problems with erratic small-signal simulation results.

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.