Amplitude and phase correction by a known transfer function
in [Matlab]
|
From: Daniel Lander on 31 May 2010 11:20 Hi, I am working on a small project which involves correcting distortions introduced by an acquisition system. The main problem is a resonate effect caused by pvc tubing which carries pressure signals to a transducer. These tubes introduce amplitude and phase distortions which can be seen within the signal in the time domain. My goal is to correct this signal for both amplitude and phase distortions. I have measured the transfer function of the tubing system and used the function 'tfestimate' to create a idealized estimate of the transfer function. The transfer function obviously has real and imaginary components (respectfully - amplitude and phase) which characterize the what happens to the signal as it passes through the system. The idea is, a signal from a new acquisition can be measured, transferred into the frequency domain, corrected by the 'known' distortion (dividing by the transfer function) and converted back to the time domain, in-phase and of the correct magnitude. The magnitude correction seems logical (the real parts of the signal / the real part of the transfer function) however where I become confused is the correction of the phase. I am not sure of the operation required by the imaginary parts of the signal and transfer function that would bring the signal back to its correct phase across the spectrum of frequencies. When I ./ the spectra of the signal by the transfer function (both parts) nothing seems to happen to the phase once back in the time domain. That is, the magnitude of the signal is corrected, however, the peaks and troughs are not aligned. If anyone could be of help here I would greatly appreciate. Regards, Daniel
|
Pages: 1 Prev: Differentiate multiple plots in same figure. Next: mean of unique values |