Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Maximum frequency, not bandwidth?

  1. Apr 10, 2013 #1

    In many places on the Internet you can read how the Nyquist frequency must be at least as high as the signals bandwidth in order to properly reconstruct the signal. Isn't this a big misnomer, since it is not the bandwidth, but the highest frequency of the signal that determines the necessary sampling rate? Why would then the word "bandwidth" still be used in this context? :smile:
  2. jcsd
  3. Apr 10, 2013 #2
    Try to make a drawing in the frequency domain of a signal with a spectral content from 40-50MHz that you sample at 20MHz. You should see there is no aliasing.
    If you want to reconstruct the original signal you will of course need extra information about the centre frequency of the original signal.

    So if you want to reconstruct the original signal without extra available information, you need to sample at 2x the maximum frequency. If you have some extra information, such as the centre frequency, 2x the bandwidth will do.
  4. Apr 10, 2013 #3
    Thanks for your answer :biggrin:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook