Ripple of S parameter
Altair EmployeeI built a model of two antennas using a wire line to transfer wireless signal and there are two ports using to calculate S parameters. Below is the result of S21. But I don't understand why there are 1dB ripples. As I think there shouldn't be that strong ripple and should be almost a single line. Can someone help me to find out if I made any mistake.
<?xml version="1.0" encoding="UTF-8"?>
Answers
-
Hi @yutourc
I have had a look at the simulation and everything seems to have been set up correctly.
I am not familiar with this specific antenna, but I see that the cylinder is quite long (0.5m). The ripples are also visible in S11 - are you sure that these are not multiple harmonic resonances from the cylinder as the frequency is increased?
Do you have a paper or measurements that you are comparing the simulation too for the expected results?
0 -
Hi Yutourc
I agree with Peter. The S11 ripple is causing the ripple in the S21. Comparing S11 and S21, where the S11 peaks, the S21 is deepest on the graph.
To get the transmitted power without mismatch, you would have to normalise S21 by 1-S11^2.
Usually, I would also recommend extending the port somewhat with an additional cylindrical section and use finer meshing there. I does not make a difference in this specific case though.
I see you used MLFMM. You can use MoM with 1 plane of geometry symmetry and 2 planes of magnetic symmetry. This then only takes 75 MB. I attach my model.
Also the 'low' setting for convergence accuracy is sufficient here. This reduces the number of required frequency samples for the interpolation.
<?xml version="1.0" encoding="UTF-8"?>
0
