Waveguide dimensions

[VegardMidt]

Hello.

I am trying to simulate a waveguide using HFSS. The dimensions are calculated using the standard equations for TE10 mode.
In the first model, I used a cut-off frequency of 2 GHz and a center frequency of 4 GHz (waveguide_2GHz.rar). The simulations seams to work, and I get a good s11 between 3.2 GHz and 4.2 GHz.

In the next simulaition, I use a cut-off frequency of 4 GHz and a center frequency of 6 GHz (waveguide_4GHz_6GHz.rar). I changed the dimentions in the model, but now i does not work. See waveguide_4GHZ.png.

Is there anything wrong with the simulation setup, or is it the model it self? View attachment waveguide_4GHz_6GHz.rarView attachment waveguide_2GHz.rar

 

[PlanarMetamaterials]

I'm not sure this has anything to do with your waveguide sizing; the waveguides seem OK. I would suspect it is your coaxial pin matching that is the problem.

You should simulate the scattering parameters of the waveguides on their own though, just to be sure. Also, I would extend the radiation boundary on the back side outwards so that it doesn't touch the PEC.

Good Luck.

 

[VegardMidt]

Thank you for your reply.

I have tried to simulate with a larger boundary on the back side, but it didn't not work. I have aldo tried to sweep the cavity depth, and it does work for some of the values. However, accoring to the equations for standard waveguide, it is given that the depth of the cavity should be lambda/4 (which is the value used for the plot in the first thread). Using the adjusted value would work, but I don't know why?

 

[nomigoraya]

Are you sure it is lambda/4 as if i remember correctly it was lambda_g/4 where lambda_g is guided wave length.
One last thing if simulation is not working fine it means for sure we are assuming something and doing one thing there must always be some issue even if it our understanding of problem or understanding of software

 

[VegardMidt]

I have used lambda_g/4. I forgot typing it. I tried to simulate a shaped probe insted of simple coaxial probe, and now it works fine. I searched for some litterature and it has something to do with the capacitive reactance from the probe:
https://www.radartutorial.eu/03.linetheory/tl11.en.html