FEKO Near Field Coupling Analysis

Faruk_Gungor

Hi,

Is it possible to perform coupling analysis with FEKO using near field data? We need to define and analyze the receiver and transmitter using near field data. Could you show an example study?

Best regards,

Faruk

Zeina

Hi @Faruk_Gungor,

You can perform coupling analysis using equivalent sources to model the Tx and Rx. Here is an example where this is performed between two horn antennas each modeled using an equivalent far field source. If you want to use near field equivalent representation, you can choose a near field instead of a far field in the step where you create your equivalent source.

To define the Tx antenna using equivalent near field, check here for equivalent source definition possibilities. To define a near field Rx, you can also check here

Hope this helps!

Best,

Zeina

Faruk_Gungor

Hi,

To compare full model coupling and near field coupling, i used two same dipole antenna from component library and I put a certain distance between two antennas and looked at the s21 value. Then, when I converted these antennas to near field format and looked at them that way, I did not get exactly the same result. Do you have any comments on what could be the reason for this? I have shared the results and the steps I took below in the attachment.

Best Regards,

Faruk

Zeina

Hi @Faruk_Gungor ,

One thing you should pay attention to when using equivalent near fields is that you must not have any geometry part "inside" the near field definition area. For instance if you have a NF equivalent source and a geometry you should make sure that the geometry is outside the NF definition space. Otherwise your equivalent source would not account correctly for coupling with the geometry parts that are within its definition area.

Having said that, I see from the second screenshot of your CADFEKO model, that your Rx definition sphere is intersecting with your source sphere. Please control this in PostFeko and double check it it is the case. This should be avoided, since it is similar to the case I described above.

Another thing you should be sure of, is the orientation of your Rx NF definition, you should change your work-plane to point towards the source (if you want them pointing to each other), so that it is correctly oriented. Now if the NF is symmetric, this might not have an impact on the results, bu always make sure you have the correct orientation in any case, as below:

These are the points that come to my mind that you should check for.

If you still have issues, please share with me the three models you have created so that I check them out: model with one dipole from which you extracted the NF data, the model with the full two dipoles, the model with the equivalent dipoles using NF equivalence.

I will be off for the holidays, so you might only hear from me early January.

Best Regards,

Zeina