FDTD S parameters - negative sign of S21
时间:03-29
整理:3721RD
点击:
Hello,
I'm trying to calculate scattering parameters from a 3D FDTD code. I consider an "abritrary shaped" scatterer within a smooth waveguide. In order to calculate the S parameters, I express the fields on the basis of modal eigen fields at two different ports as a sum of a forward and a backward component, for each mode (E=B+F, backward and forward waves). I excite the waveguide with a gaussian pulse, with temporal legnth equal to 6 times the temporal standard deviation \sigma_t. The same time is used for the simulation. I calculate the S21 as Forward_2/Forward_1 but I receive negative values (in dB). I expect to get negative values but close to zero, as a lot of the power will bypass the scatterer. In the mean time, S11 is still low, meaning that there is no significant reflection. What can it happen? Is it possible that the time duration is inadequate and I need more time in order to identify possible resonance phenomena?
Thank you in advance.
ps: Scatterer is an abrupt change in the dimensions of the waveguide, creating a staircase.
I'm trying to calculate scattering parameters from a 3D FDTD code. I consider an "abritrary shaped" scatterer within a smooth waveguide. In order to calculate the S parameters, I express the fields on the basis of modal eigen fields at two different ports as a sum of a forward and a backward component, for each mode (E=B+F, backward and forward waves). I excite the waveguide with a gaussian pulse, with temporal legnth equal to 6 times the temporal standard deviation \sigma_t. The same time is used for the simulation. I calculate the S21 as Forward_2/Forward_1 but I receive negative values (in dB). I expect to get negative values but close to zero, as a lot of the power will bypass the scatterer. In the mean time, S11 is still low, meaning that there is no significant reflection. What can it happen? Is it possible that the time duration is inadequate and I need more time in order to identify possible resonance phenomena?
Thank you in advance.
ps: Scatterer is an abrupt change in the dimensions of the waveguide, creating a staircase.
If you have a lossless DUT, you should have |S11|^2 + |S21|^2 = 1
so maybe check that first.
In my case this is not valid. Note that I excite with TE11 and measure the S11 and S21 of the same mode. I have simulated in CST studio a holow waveguide with a "brick" inside, and measure S11 and S21 and |S11|^2 + |S21|^2 = 1 is not the case. I suspect mode conversion due to the obstacle
栏目分类