HFSS15: Assigning Lumped Ports for Terminal Solutions

1. Select the object face to which you want to assign the port.

Each terminal that is identified by an edge selection must have each edge contained by some non-port face.

2. Click HFSS> or HFSS-IE>Excitations>Assign>Lumped Port.

3. Define the complex Full Port Impedance of the port in the Resistance and the Reactance text boxes.

You can assign a variable as these values. This variable can be dependent on the frequency, which allows use of a dataset for frequency dependent impedance.

4. Click Next or select the Post Processing tab.

This displays the Lumped Port: Post Processing window. Values here affect S-Parameters only. Port processing operations do not affect field plots. By default, the port renormalizing impedance for all terminals is 50 Ohms.

If you want to enter a complex impedance, enter it in the following form:

<re> + <im>j

5. To specify a different full port impedance, type a value in the Full Port Impedance text box. Select the corresponding unit in the drop down menu.

6. For HFSS single frequency solves, adaptive pass solutions, interpolating sweeps, and fast sweeps, if the port has a single rectangular face and the terminal assigned to the lumped port has either a single edge or a single face, you have the option to check the Deembed.

The Deembed settings do not appear for lumped ports for transient or transient network analysis designs

Checking Deembed enables an analytic formula for the parasitic lump inductance of the port geometry resulting in a frequency-dependent calibration or "de-embedding" of lumped ports, which can be thought of as "deembedding" the lumped port structure. Lumped port calibration is a post processing step, using port geometric parameters that are extracted and saved when solving. The calibration is also available for solution results produced prior to HFSS version 14, in which case the port geometric parameters are extracted and saved once per design variation at the time of the first matrix data load for it.

After possible interpolation, renormalization, and deembedding applied to the matrix data for each frequency, the calibration formulas modify the Z matrix and then compute a replacement for the post processed S and Y matrices. The same calibration applies when calculating adaptive convergence, since adaptive convergence is based on fully post processed matrices. The frequency dependent formulas are encapsulated within the loaded solution and used when it supplies frequency data. Solutions sent over to Designer via the dynamic link have the same support, based on exactly those lumped ports which were selected for this option.

The Deembed option can be turned on or off as a post processing operation without invalidating solutions. If multiple lumped ports are selected in the Excitations List panel, changes to the common Deembed property change the setting for all lumped ports at once, but only for those that support calibration.

If a port is selected for calibration but the validation checks fail for some variation (which could happen with a parametric solve), then the matrix data will not load for that variation. Also, if the fields are to include port post processing effects, then the fields do not load.