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.

Using Z-Parameters to find input and output impedance

Status
Not open for further replies.

Hawaslsh

Full Member level 3
Joined
Mar 13, 2015
Messages
167
Helped
5
Reputation
10
Reaction score
7
Trophy points
1,298
Location
Washington DC, USA
Activity points
3,461
Hello all,

I made some 2-port measurements using a a VNA and needed to deembed a length of transmission line. So i converted the S-parameters to ABCD, did the matrix math to remove the t-line then converted back to S-parameters.

I need to know know the input and output impedance so i converted the deembedded S-parameters to Z-parameters. For the longest time I thought Z11 and Z22 would represent my input and output impedance but I ran across:
Capture.PNG

Have I been interpreting Z11 and Z22 wrong all these years? Should I be using the equations above to solve for the input and output impedance?

Side question: If i had a one port network would Z11 be in the input impedance?

Bonus question: I just realized the method I am doing only deembedds a length of transmission line from port 1. I want to remove an identical length of line from port 2 but my matrix math is a bit rusty. Using ABCD matrices, the initial condition is [tline][tline]=[measurement]. I can perform [tline]=[tline]^-1*[measurement] to remove the tline from port 1, but how do I go about getting rid of the second tline?

Thanks in advance!
 
Last edited:

Znn is, by definition, the impedance seen looking into port n when the current at all other ports is equal to zero (open circuit). So if the other ports are open, then Z11 is your input impedance. For a one port network, Z11 will always be your input impedance.
 

This might relate to a transformer with impedance ratios that are defined by coupling factors to load ZL, but does not look right to me without coefficients of coupling.
 

Hello all,

I made some 2-port measurements using a a VNA and needed to deembed a length of transmission line. So i converted the S-parameters to ABCD, did the matrix math to remove the t-line then converted back to S-parameters.

I need to know know the input and output impedance so i converted the deembedded S-parameters to Z-parameters. For the longest time I thought Z11 and Z22 would represent my input and output impedance but I ran across:
View attachment 119036

Have I been interpreting Z11 and Z22 wrong all these years? Should I be using the equations above to solve for the input and output impedance?

Side question: If i had a one port network would Z11 be in the input impedance?

Bonus question: I just realized the method I am doing only deembedds a length of transmission line from port 1. I want to remove an identical length of line from port 2 but my matrix math is a bit rusty. Using ABCD matrices, the initial condition is [tline][tline]=[measurement]. I can perform [tline]=[tline]^-1*[measurement] to remove the tline from port 1, but how do I go about getting rid of the second tline?

Thanks in advance!


1- When you work with a two-port network, Z11 is NOT input impedance, Z22 is also NOT output impedance. You have to use the equations which you shared.

2- In a one-port network, Z11 is input impedance.

3- Sorry, I have no experience about this.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top