Thanks Andrew! Problem solved. I am using Cadence version ic123_isr10. spectre-W says W is an unbound variable, so I could not get the spectre version directly. However, from the simulation log file, I see that it is loading packages from mmsim151_isr11, so I assume the spectre version is 15.1 Indeed, PSS support s-parameter if I choose harmonic balance engine and spline for interp. However, if I choose shooting engine, it gives the following error: Error found by spectre during periodic steady state analysis `pss'. ERROR (SPCRTRF-15160): Distributed Components are not supported directly for Autonomous PSS or Envelop Analysis with Shooting engine. 1) Firstly, it's recommend to use Harmonic Balance engine for Distributed Components; 2) Or else, for this case and with Shooting engine, try the Rational Fitting Technique(set interp=rational) for NPORT components. Actually the component that has the s-parameter file comes from a third party tool, which chooses 'spline' for interp . I tried to change it to 'rational' following the suggestions in the error message, simulation runs but did not converge. After your suggestion, I used bbspice, then it runs and results seem to be ok. Thanks for the help!
↧