N
Member
Conclusion first, reasoning after.
1. Non-causal S-parameters produce a response before the excitation, which is physically impossible.
2. Non-passive data can make a transient simulation diverge or produce spurious gain.
3. Both issues typically come from measurement noise or from extrapolating outside the measured band.
4. Run a causality and passivity check before using any measured block in a time-domain simulation.
Happy to hear a better approach.
1. Non-causal S-parameters produce a response before the excitation, which is physically impossible.
2. Non-passive data can make a transient simulation diverge or produce spurious gain.
3. Both issues typically come from measurement noise or from extrapolating outside the measured band.
4. Run a causality and passivity check before using any measured block in a time-domain simulation.
Happy to hear a better approach.