Transmission loss ( S_21,dB ), transient analysis the same?
I have the transmission loss ( S_21,dB ) of a simulation and I wanted to confirm, if I would get the same result for transient analysis.
But the results differed.
I appreciate your help.
Hello,
I don't know what you are doing, but when performing a transient analysis, the stimulus must be narrow band as your S21 measurement is for one frequency only. You can have a wide band stimulus, but in that case you need to convert to frequency domain to get useful frequency domain data.
Are you doing the frequency and time domain analysis in the same program?
Hello,
thank you for the fast response.
I have data for S21 from 0GHz to 50GHz.
Yes, the analysis is done in the same program.
Hello,
when doing your time domain simulation, what stimulus do you use? Are you driving from the same reference impedance as you did with the frequency domain simulation?
Providing some additional info enables us to give a better answer.
I use a sin as stimulus and the reference impedances are identical.
The sin has amplitude 1V and the frequency is adjustable.
First I used a frequency of 10GHz for the sin ( for time domain analysis ) and checked the result with S21 at the same frequency 10GHz. But the results differed.
Do I need a reference impedance at the input, too ( for time domain analysis ) ?
Because for the transient analysis, I only use a reference impedance at the output.
I′m not so familiar with the transient analysis.
Hello,
when you used 50 Ohms to derive the S-parameters, you also should use 50 Ohms in series with your input voltage source.
When you take the EMF of your source 2V, the output of your 2-port equals S21.
The S-Parameters should be right.
Do I need 50 Ohms in series for transient analysis with the input voltage souce, too?
Hello,
The whole S-parameter system is based on voltages with respect to a termination for maximum power transfer.
So when running a transient simulation based on lumped component base (voltage sources, current sources, resistors, etc), you need to add 50 Ohms in series with your voltage source to get meaningful results.
By taking the voltage source's EMF = 2V, you have an incident wave of 1V. The actual voltage you "measure" on the output of your voltage source with 50 Ohms in series, is the sum of the forward wave and reflected wave by the two-port.
You can see that when you left our new source (voltage source + 50 Ohms) open. You will "measure" 2V. This is not strange as an open has a RC = +1, so when the forward voltage is 1V, the reverse voltage is also 1V, so actual voltage at input of two (=output of source) = 2V.
when using a "port" in a microwave design program, the port's impedance is embedded inside the port.