[Moved] conversion from frequency domain to time domain using matlab

[sajjadwazir]

This is the file in touchstone(S1P) format . It is in frequency
domain .we need to convert it to time domain.we got it from Vector
Network analyzer.And frequency rang is from 3 to 10 GHz. Total number of
sample is 201.the S11 is in dB.the file contain three column first column
contain frequency second column dBS11 and third show angleS11
(# frequency_unit S DB R impedance
freq dbS11 angS11)
(i think we did not need to convert angS11) just first and second column
need to convert.using mat-lab please can any ona can help me

 

[KlausST]

Hi,

Mathematically an inverse fourier transform should work.

Klaus

 

[vGoodtimes]

What is the intent? For example, if I gave you a picture of the result and it looked almost like noise what would you think?

I say this because unless you have a near 0 S11 (not 0db...) across the entire bandwidth, the result of a transform back to time domain would look mostly like noise.

(also, angle is among the most important aspects. It plays a large role in determining shape as it primarily affects peaks and edges -- the things you can actually see in time domain plots.)

 

[c_mitra]

If you have a plot of a voltage signal as a function of time (how the voltage changes with time), the function is in time domain. You need to use Fourier transformation to get this into the frequency domain.

When you do the Fourier transformation, you will get a complex result. Although the original voltage function was real, the Fourier transform is complex. Mostly we use the modulus of the complex quantity.

The x-axis after the Fourier transformation becomes frequency scale (it is now proportional to 1/t) and is often called the power spectrum. Another Fourier transformation will get back the original voltage versus time graph.

Not quite so; if you use the modulus of the complex quantity, you will now get the voltage that is complex now. If you convert the complex number to the polar form (amplitude and phase), you will get another result. All are related but are often difficult to figure out just by looking at it.

Transformation of log quantities are another matter altogether.