50 Ohm Inactive Port Termination HFSS

[windscare2]

Simulating using Terminal Driven Mode

Comparing 2 simulations:
1) Port 1- Lumped Port Element (50 ohm) to Microstrip terminated with P2 - Lumped RLC resistance set to 50 ohm
Z Parameters are measured to verify microstrip line (width = 3mm) is properly matched to 50 Ohms in FR4 substrate.

2) Port 1 - Lumped but now Port 2 is also set to a Lumped Port (renormalized to 50 Ohm) instead of Lumped RLC
Z Parameters show that the circuit is consistent with a short to GND, with the microstrip line behaving as quarter-wave transformer at f=800MHz (Z-> inf)

Figure 1: Microstrip set-up with Port 1 Lumped Element, terminated with 50 Ohm Lumped RLC

Figure 2: Z Parameter Plot of Simulation #1 with P2 - Lumped RLC

Figure 3: Z Parameter Plot of Simulation #2 with Port 2 - Lumped Element


Without using Wave Ports/Modal Solution Type, how can I set my Lumped Element Port 2 to default 50 Ohm termination when inactive.

Thank you!

 

[PlanarMetamaterials]

Welcome, windscare2,

It sounds like something is wrong with the configuration of the lumped port in the second simulation. Can you show some additional details on its setup? In particular, I am wondering if the voltage/integration line is set properly.

 

[windscare2]

Welcome, windscare2,

It sounds like something is wrong with the configuration of the lumped port in the second simulation. Can you show some additional details on its setup? In particular, I am wondering if the voltage/integration line is set properly.

Thank you for the reply!
I tried what you recommended and changed my Solution Type to Modal for the integration line, here's my Lumped Port Element for both Terminal & Modal Solution Types.
Z Parameter plot shows re(Z(1,1)) for Modal Solution plot (compared to re(Zt(Trace1_T1,Trace1_T1)) for Terminal Solution

]

 

[PlanarMetamaterials]

Yes, that all appears to be correct - interesting.
Can you plot the fields travelling along the strip and the magnitude of E/J on the ground plane? I wonder if there's some setup glitch that might have been overlooked.

 

[windscare2]

Yes, that all appears to be correct - interesting.
Can you plot the fields travelling along the strip and the magnitude of E/J on the ground plane? I wonder if there's some setup glitch that might have been overlooked.

Nice idea!

Heres the surface currents at 600MHz, when the P2 is terminated to 50 Ohm (Simulation 1)

Heres the same at 600 MHz when P2 is Lumped Element (Simulation 2) (phase is off by 10 degrees)

The Sources are set up so that the Active Port 1 is feeding into Inactive Port 2

Even including Post Processing/Terminating P2 manually from this menu hasn't helped

 

[PlanarMetamaterials]

Ok, your simulation is fine. I overlooked an important detail.

Z11 is only equivalent to input impedance when there is only one port. You can see this in the definition of Z-parameters; it involves assuming that the other ports are open-circuited, not matched. When you add a second port, obviously Z11 is going to change!

You can confirm the ports are both well-matched to the line by observing S11, S12, and S22. Alternatively, you can derive the expression for Z11 of a two-port network, and confirm that it matches what HFSS is showing you.

Good luck!

 

[windscare2]

Ok, your simulation is fine. I overlooked an important detail.

Z11 is only equivalent to input impedance when there is only one port. You can see this in the definition of Z-parameters; it involves assuming that the other ports are open-circuited, not matched. When you add a second port, obviously Z11 is going to change!

You can confirm the ports are both well-matched to the line by observing S11, S12, and S22. Alternatively, you can derive the expression for Z11 of a two-port network, and confirm that it matches what HFSS is showing you.

Good luck!

Wow, thank you I completely overlooked that!

Another way we can use to calculate the impedance of the line was to take the Z11 of the Short Circuit Port 2(Zsc) and the Open Circuit Port 2(Zoc), and to take the sqrt of those together-> Zo = sqrt(Zsc*Zoc) and that amazingly came up to 50 Ohms in my simulation!

Thank you very much for your time, I appreciate the help!