Interpreting far-field E filed due to a time signal specified in POSTFEKO

Michael R

To convey my concerns, I will present a special case:

1. I do a FEKO simulation based on a signal with a frequency f.
2. In POSTFEKO, I then specify a time signal f(f) = sin(2*pi*f*t)
3. In POSTFEKO, I then view the far field E field and see that it is non-zero beginning at t = 0 s. With this, it is not near zero such that one may think the nonzero value is just a numerical artifact. Instead, it is a large nonzero value.

The behavior mentioned in (3) does not make sense to me because I think that if the input signal is zero at t = 0, then the Far field E field should also be zero at t = 0. Can someone please provide insight on what is going on? It would be very much appreciated.

Jaehoon

The short answer is that the signal repeats both forwards and backwards in time due to the nature of the Fourier transform implementation in Feko. So, the signal does begin at time=0 and is not zero a time=0.

To get the exact time response, the range of the simulation frequency in CADFEKO should be defined with the following Fourier transform-based settings.

    1) Minimum frequency = 1/(Signal duration)

    2) Maximum frequency = 1/2 * 1/(Number of samples)

Thank you.