Re: Phase plot problems

I'm calculating the transfer function between input and
output signals. My inputs to the system were Gaussian white
noise. Therefore, the magnitude and phase plots I obtain
were very noise (even with a large number of samples).
How are you calculating your transfer function? That fact that you use
noise has nothing to do with obtaining a noisy transfer function.
Injection (white/colored) noise into a system is actually a pretty
common way to measure transfer functions. Have a look at the function
'tfestimate' and play with the number of averages. By increasing the
number of averages you can get a very accurate transfer function
(given enough coherence, stationary of you system, ...).
Also, I'm using FFT averaging to reduce the noise
variation.
See tfestimate mentioned before. Don't write you own code, it takes
some effort to get all the details right.

This is an issue for the phase plot, because even though I
can barely see a general trend in the plot, there are a
lot of noise samples between +pi and -pi through out the
graph. I can't even use phase unwrapping because of the
large jump in angles.

If the lines jumping from +-pi are the problem are the problem why not
simply plot the points (e.g. semilogx(f,phi,'.'). If you insist on
using a solid line, you might use this function:

function [tgap,pgap] = phasegap(t,p)
%function [tgap,pgap] = phasegap(t,p)
%inserts a NaN where diff(p) > pi in both t and p to create a gap in
the plot:
%[tg,pg] = phasegap(t,p); semilogx(tg,pg)

jumpIndex = find(abs(diff(p))>pi);
jumpIndex = [0;jumpIndex;length(jumpIndex)];
pgap = NaN(length(p)+length(jumpIndex),1);
tgap = pgap;
for k = 1:length(jumpIndex)-1;
tgap(jumpIndex(k)+1+k-1:jumpIndex(k+1)+k-1) = t(jumpIndex(k)
+1:jumpIndex(k+1));
pgap(jumpIndex(k)+1+k-1:jumpIndex(k+1)+k-1) = p(jumpIndex(k)
+1:jumpIndex(k+1));
end

If anyone nows how to vectorize the for-loop, I would be glad to hear
it.

HTH,
Bas
.

Relevant Pages

  • Re: Good enough for crypto?
    ... be considered the actual entropy source. ... resistor as a thermal noise source. ... perfect gaussian noise source with extremely high bandwidth. ... amplifier, which has its own transfer function, and filter are added. ...
    (sci.crypt)
  • Re: Transfer Function Estimation - Averaging?
    ... I have a question about transfer function estimation relating ... The data that I have is daily records of measurement of earth's noise. ... the corresponding part in frequency domain and then added them together ... and averaging in frequency domain with same amount of segmenting and ...
    (comp.dsp)
  • Re: Data smoothing algorithms? - Thank you all
    ... noise being conspicuously distinct from the signal. ... If I plot the data and look at the plot from a distance far enough to ... A trample path is a line through a landscape from one place ... Interpolation would be artificially increasing the number of data ...
    (comp.lang.python)
  • Re: Mic2496 input noise (was: Mic level A/D Converters?)
    ... give about 2.62 uV RMS inputnoisefor theMic2496. ... size and function that Len used to do the plot. ... Assuming the noise is white and is at a level L in the spectrum plot, ...
    (rec.audio.pro)
  • Mic2496 input noise (was: Mic level A/D Converters?)
    ... give about 2.62 uV RMS inputnoisefor theMic2496. ... it is an invalid procedure unless you use the same FFT window ... size and function that Len used to do the plot. ... Assuming the noise is white and is at a level L in the spectrum plot, ...
    (rec.audio.pro)
Loading