CST excitation function (theoretical)...

Hello everyone,

My question is mostly theoretical, regarding the CST excistation function :

How does CST (or similar softwares) excite the model during a frequency sweep?
Assuming my purpose is S-Parameters of the model, why is it excited with a Rectangular pulse or Gaussian, and not with a Continuous Sine Wave ?

Thanks alot for your attention,

Pushhead.

Pushhead said:

Hello everyone,

My question is mostly theoretical, regarding the CST excistation function :

How does CST (or similar softwares) excite the model during a frequency sweep?
Assuming my purpose is S-Parameters of the model, why is it excited with a Rectangular pulse or Gaussian, and not with a Continuous Sine Wave ?

Thanks alot for your attention,

Pushhead.

Click to expand...

If we assume we're in time domain...
Structure is feeded at the port with gaussian pulse (modal fields are modulated by this pulse). This gaussian pulse have defined spectrum according to highest frequency range.
This excitation signal is called "i" in MWS/time signals (during simulation run)
Software also monitores signal reflected from the structure back to port; this is called "o".
Excitation and reflected signal are then transformed to frequency domain using FFT and frequency response (S parameters) are then calculated as division of transformed signals.

You imagine excitation only with sine wave. Sine wave contains only one frequency, thus you'll get S-parameters only at one frequency point.

Another choice is bandlimited signal for structures with cut-off frequency.
Then sine modulated with Gaussian is used. This signal has spectrum spreaded around center frequency according to freq. of sine wave.

Cheers

e.

Great , 10x alot.

That makes sense ( What I meant was a sweep using a CW , changing the sine freq. each iteration, but nm, you really helped me) .

Another one if I may :

Consider a signal generated at the ports of a charging/discharging capacitor, limited in time from 0 to a few micro-seconds.

If I understand your explanation right, then I have to estimate the freq. bandwidth of the signal using FFT or (1/Tmax), right ?
(and then use it in the Frequency Range dialog box).

Thanks, P.