For the past couple of weeks I am trying to simulate an electrically small antenna (dipole like) to extract impedances for a matching circuit. The only issue that I have run into is that for the low frequencies the the S-parameter does not stabilise < 0. I have tried everything, increasing the mesh, increasing the boundaries, increasing the bandwidth of the signal. The simulation time is up to 10h but still no change.
The bandwidth is 10 MHz - 4 GHz, f0 = (f_start + f_stop)/2, and fc = f0/2. I tried both MUR and PLM_20 as boundaries. Fine mesh of 0.15mm, coarse mesh of 1 mm, and a simulation box of +- lambda. The dipole is excited by a 50 ohms lumped port, which is 2x2x0.035mm in size (gap between the dipoles). I tried to increase fc (bandwidth) but the energy levels remain the same at circa -7dB... I can confirm that from 600 MHz to 4 GHz the result is correct. I did try to simulate only the 10 MHz to 600 MHz part with the same result. I have read the FDTD theory and I don't know if I missed something (https://en.wikipedia.org/wiki/Finite-di ... ain_method).
Also, to reduce the simulation time I was wondering if I could use symmetric planes? I did a quick search on the forums but I didn't find any one explaining how, so please correct me on what I am doing wrong.
For symmetries XZ: Electric; YZ: Magnetic; XY: Open;
BC = {'PEC' 'MUR' 'PMC' 'MUR' 'MUR' 'MUR'}; % boundary
Then the mesh limits should be something like;
mesh.x = [0, ... , SimBox(1)];
mesh.y = [0, ... , SimBox(2)];
mesh.z = [-SimBox(3), ... , SimBox(3)];