Learning Forum

[Farooqui]

I am simulating a narrow band antenna (0.5 % bandwidth) designed to work at 120 MHz. It is a monopole antenna. The simualted S-parameters (S11) showing the resonance changes by few MHz every time I increase the size of the radiation box. The radiation box is now multiple wavelengths away from the antenna and even then it is affecting the simulated S-parameters. Any solution to this problem?

[Praneeth Munaga]

The radiation boundary is the simplest form of open boundary condition. Ideally it should be placed at least quarter wavelength from the antenna.
Can you share a comparison result of your S parameters for different distances as you said?
Meanwhile you can try either PML or FE-BI boundary.
I would like to know more details about your radiation boundary setup to help us serve you better.
Did you define radiation boundary in all the directions?
Did you define uniform padding in all the directions?
Best regards.

[Farooqui]

Thanks for the response. I defined radiation boundary in all the direction- I enclosed my antenna in a radiation box. How do you define padding in radiation boundary. I did not define any padding.

[Praneeth Munaga]

Air padding or simply padding is the distance between the antenna and the radiation boundary periphery.
In your first question you have mentioned that you have more distance between antenna and radiation boundary. May I know how did you evaluate this.
Please go through the HFSS help documentation for detailed information on auto open region in HFSS.
All the very best.

[Farooqui]

As I mentioned before. the distance between my antenna and the radiation boundary is more than quarter wavelength at the frequency of simulation (120 MHz) in all directions, and I kept on increasing this distance to see when it stops affecting the simulated results. But it keeps on affecting the resutls (shift in resonance) even at a distance of multiple wavelengths.

[Praneeth Munaga]

Please share screen shots of the results comparison along with calculated distance between antenna and radiation boundary.
As mentioned before you can also use PML and FE-BI open boundary conditions and see the effect.
Best regards.

[AndyJP]

it may happen that you just defined a bad convergence in the setup. make convergence delta 10 times smaller than the default, and set at least 14 iterations limit with at least 2 consequent converged solutions.
Use meshing at the frequency of interest just above the resonance, or use 2 meshing frequencies above and below. I hope you did not set the dielectric constant for the air/vacuum less than 1?

[AndyJP]

P.S. the padding should works better at odd multiple of quarter-wavelengths. It works like an impedance transformer. And the broadside should be more or less parallel to the radiation boundary. I.e. the radiated TEM wave should hit the wall flat.