Does a PNA show gain across a band using FFT?

  • Thread starter Thread starter lennybogzy
  • Start date Start date
  • Tags Tags
    Band Fft Gain
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
Sign up Log in
6 replies · 2K views
lennybogzy
Messages
94
Reaction score
0
Does a PNA show FFT

If I'm looking at gain across a band on a PNA could one say that I'm looking at an FFT or is that complete nonsense?
 
Last edited:

Discussion

Engineering news on Phys.org
Older instruments performed a frequency sweep and recorded the response at each frequency. To use an FFT you would need to excite with a delta function, measure the impulse response, and perform the transform. I don't think that even modern PNA's do that...
 
Last edited:
No, that is not how a VNA works (PNA is an Agilent acronym, I have no idea what the P stands for). A VNA basically performs a frequency sweep using a syntheziser and then records the response using a mixer where the RF comes from the DUT ans the LO comes from the syntheziser (there are actually several mixers involved in the detection, since a VNA measures the vector response).

There are FFT analyzers out there (and they are called FFT analysers, do distinguish them from spectrum analyzers, scalar analyzers and VNAs), but they are limited to relatively low frequencies (a few hundred kHz, the ones I use go upp to 100 kHz).

There are a few good Application notes on the Agilient website which described how their PNAs work in some detail.
 
Being interviewed for a position: the guy interviewing me cut me off with something like "so you were looking at an FFT" (I was talking about gain in k-band). I started to correct him but he was adamant about me "looking at an FFT" so I kinda let it go.
 
Well you are looking at an FT it's not an FFT, but people like the word FFT too much these days.
 
Come again? FT - Fourier transform? What do you mean?

My understanding is that I'm looking at something directly measured, rather than 'computed' in that sense
 
Last edited:
You are right, butl the Fourier transform does have physical reality in a lot of cases. If you add two batteries in series the output voltage is the sum of the voltages. You would still call it a sum even if it is not computed.
A frequency vs. power spectrum represents the absolute value squared of the signal in frequency space. There are examples of naturally occurring Fourier transforms for example in optics.
The FFT on the other hand is actually a "Fast Fourier Transform" a special algorithm to get a frequency spectrum. But people mix their words all the time and call any Fourier transform an FFT even if a DCT (a similar algorithm) was calculated or if the spectrum does not have a computational origin at all.