I have been acquiring data through ADC and into memory, the input signal are square waves that step down in frequency as time goes by. Due to memory, I have a fixed maximum number of samples per FFT. But I would like to compare their magnitudes, and phase along side each other. The max length of FFT is good enough for frequency resolution, I am just wondering if I have a bunch of data that's been processed through FFT, can I sort of just add them to combine their total magnitude and phase? or will the addition cause bias and distortion?

Also, I was wondering since I noticed that at low frequencies, the flicker noise seems to increase the noise floor, is it possible to have a magnitude correction filter or equalizer to push the rest of the signals up such that I can have a common noise floor when I go through all my input signals?

Thank you.

  • $\begingroup$ Extra note to my question is, if I use a sliding DFT, with a sufficiently small window and kind of look at the data in a 3D plot, will I be able to compare the magnitude that way? if so, how should I gauge the size of the window for each DFT, and how much do I slide per FFT? (input signals range from 0.5 Hz to 30 KHz, Fs = 128KHz $\endgroup$ – user8481 Jan 28 '15 at 16:22
  • $\begingroup$ Adding the DFTs will result in the same thing as adding the signals before taking the DFT since it is a linear transform. $\endgroup$ – Oscar Jan 30 '15 at 10:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.