Loading ...
Sorry, an error occurred while loading the content.

Cross-correlation measurements

Expand Messages
  • Graham Mcleod
    Hi Wolf, I am interested in using SL for this function. A 20dB reduction in Noise Floor beckons. I have this defined as Vector addition of the (FFT o/p)
    Message 1 of 1 , Jul 20, 2012
    • 0 Attachment
      Hi Wolf,
      I am interested in using SL for this function. A 20dB reduction in Noise Floor beckons.
      I have this defined as "Vector addition of the (FFT o/p) samples from 2 signal paths,
      with subsequent Averaging". Correlated signals/noise have the same phase, so their
      magnitudes add, whilst Uncorrelated signals tend to cancel towards Zero.
      A clear example of this use is shown in the Rohde & Schwarz Application Note 1EF56,
      Figure 1, and the paragraph immediately underneath.
      This appears to differ from your "Correlogram", in that the output which I need is the
      Same as the two FFT outputs, ie Log Amplitude versus (Log) Frequency on the X axis.
      I wonder if this can be done from just two real signal inputs, CH1 and CH2 ??.
      Or are the "real-only" phase-angle values irretrievably thrown away in the process?.
      With separate FFT's, these are irrelevant and meaningless, but for the cross-correlation,
      presumably the "software local oscillator" angles need to be Common to both FFT's.
      Is such a feature possible in Speclab?. I'm unlikely to need it for a year or so.
      Graham G8PHA.
       
    Your message has been successfully submitted and would be delivered to recipients shortly.