questions about s-parameters

[TecHNooB]

do they account for intermediary reflections between ports or just the steady state? are there any assumptions made about where the reflections occur? do they accurately predict transient behavior when reflections are significant?

 

[futuristicmonkey]

do they account for intermediary reflections between ports or just the steady state? are there any assumptions made about where the reflections occur? do they accurately predict transient behavior when reflections are significant?

Steady state. The whole point is you can treat the device as a black box and only have to worry about the terminal characteristics.

 

[Born2bwire]

For transients they typically talk about other measurements. Eye diagrams are often used for such measurements.

While they are steady state, there are some intricacies involved about where the reflections occur in terms of the device under test (DUT). That is, the measurement device (network analyzer) is connected via transmission line to a port on the the DUT. A proper measurement takes into account the possible measured reflections of the input ports and transmission lines used. For example, the input port on the DUT may not be matched to the impedance of the DUT itself. A naive measurement would include the effects of the port mismatch and so forth. However, in the end result has the DUT itself modeled as a blackbox network of ports as futuristicmonkey stated.

 

[PsiStar]

I go back & forth between S-par & impedance profiles (tdr plot). One can be derived from the other. S-par tells you how bad in the freq domain & the impedance profile tells where the issues are ... if that is what you want to know.

S-par data sets for multi-port devices can be plugged into many simulators to see how it plays with the rest of the system.