Since flicker noise is inversely propotional to the noise frequency, the noise is infinity at DC. What I didn't understand from EE240 Elad Alon and even Razavi's book is that it requires a long time to see the infinite noise and so on. Can anybody please give me a clear explanation for this? Thanks
In order to find (i.e. measure its effect) 1Hz noise, you need at least 1 second. For 1mHz 1000 seconds, for 1µHz nearly 12 days. And for "DC noise" (0Hz) eternally.
Few people can wait so long ;-)
HTH! erikl
As you go lower in frequency the bandwidth declines. I have read one scientific paper published many decades ago that reported 1/f noise at a measurement time of days or was it years. One of my Caltech friends claims that 1/f noise exists in all areas, such as mechanics. There is a small but finite probability that the legs of the chair you are sitting on will extend (from 1/f noise) and push your head into the ceiling.