Hey all together,

I have samples recorded by an VNA in frequency domain and want to display just the spectrum of a certain time-domain event (e.g. the reflection of a DUT).

So I perform an invers FFT on my samples to get the time domain values where I zero out everything except the peak caused by the DUT. Of course, back in frequency domain this rectangular window will causes huge ripples. So I want to apply for example a Kaiser Window instead of the rectangular one to smooth the change.

My questions are:

- Why is in App. Note http://cp.literature.agilent.com/litweb ... 5723EN.pdf on p. 18 written that "the gating function is windowed in the frequency domain before being transformed to the time domain" and that one should perform a convolution (in frequency domain)?

1st: Isn't the "Window" my "gating function"?

2nd: Although the app note suggests working in frequency domain, there should be no difference between appliing the window in time domain (multiplication), or in frequency domain (convolution)?!

- As I apply another than a rectangular window the amplitude of the DUT's time domain reflection pattern is changed by some factor thus the "true spectrum" of the DUT is changed. Is this the price one has to pay for less ripples? Or is there a way to take this into account by some correction?

Regards

I have samples recorded by an VNA in frequency domain and want to display just the spectrum of a certain time-domain event (e.g. the reflection of a DUT).

So I perform an invers FFT on my samples to get the time domain values where I zero out everything except the peak caused by the DUT. Of course, back in frequency domain this rectangular window will causes huge ripples. So I want to apply for example a Kaiser Window instead of the rectangular one to smooth the change.

My questions are:

- Why is in App. Note http://cp.literature.agilent.com/litweb ... 5723EN.pdf on p. 18 written that "the gating function is windowed in the frequency domain before being transformed to the time domain" and that one should perform a convolution (in frequency domain)?

1st: Isn't the "Window" my "gating function"?

2nd: Although the app note suggests working in frequency domain, there should be no difference between appliing the window in time domain (multiplication), or in frequency domain (convolution)?!

- As I apply another than a rectangular window the amplitude of the DUT's time domain reflection pattern is changed by some factor thus the "true spectrum" of the DUT is changed. Is this the price one has to pay for less ripples? Or is there a way to take this into account by some correction?

Regards