Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

One-port to two-port s-parameter conversion

Status
Not open for further replies.

buin

Newbie
Joined
Jan 17, 2022
Messages
4
Helped
0
Reputation
0
Reaction score
1
Trophy points
3
Activity points
36
Hi there,
Suppose that I have a one-port s-parameter measurement (or simulation) data of a transmission line with one side terminated open, what is the best way to estimate its two-port s-parameters?
byoung
 

It is not possible to calculate the 2-port S-parameters from that single 1-port data.

If you are sure that your line is well matched (S11 close to zero), a rough estimate is that unterminated S11 is two times the lines insertion loss (S21)
 

Generally it's not possible. You can try to fit a transmission line model with limited number of parameters (e.g. Z0, electrical length, attenuation) to the S11 measurement.
 

It is not possible to calculate the 2-port S-parameters from that single 1-port data.

If you are sure that your line is well matched (S11 close to zero), a rough estimate is that unterminated S11 is two times the lines insertion loss (S21)
Thanks for the comment! Yes, I agree to your comment. However, recently I learned that Keysight has a software that can do the estimation. Basically, it is to provide a simple way for deembedding of EVB traces. All you need is just one-port s-parameter data with one side open. I took two-port data from one-port measurement one using the tool, and then compared it with (original) two-port measurement data, and it is not perfect, but they are reasonably well matched, which surprised me. :-o I am wondering about the principle of work.
 

Generally it's not possible. You can try to fit a transmission line model with limited number of parameters (e.g. Z0, electrical length, attenuation) to the S11 measurement.
Hi FvM, thanks for your comment. I replied to Volker's comment, which can be the same feedback to you.
 

As stated, it can work with a model based approach. You need to put in multiple s11 points to identify e.g Z0 and length. You can play around with circuit simulator and a transmission line model to see how the s11 curve depends on Z0 and length.
 

However, recently I learned that Keysight has a software that can do the estimation.
Do you have more details? I am using Keysight ADS, but don't know what estimate you refer to?

edit: Do you mean the TDR front panel that allows to calculate line impedance from reflection using time domain transformation?
 

Do you have more details? I am using Keysight ADS, but don't know what estimate you refer to?

edit: Do you mean the TDR front panel that allows to calculate line impedance from reflection using time domain transformation?
Keysight's N19301B(PLTS) with N19305B AFR option can do. You can install the demo version for one month.
 
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top