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

Noise what is it?

  1. Jan 13, 2008 #1
    Can someone please explain to me or preferibly point me torwards a resource that will explain what is ment by noise? I know that it s some kind of unwanted disturbance in a signal and that we want to get rid of it, but where does it come from. when i search online about this i find a myrid of ways of how to filter out "noise" but where does it come from?
    I also noticed that there is diffrent types of noise: Shot noise, Thermal noise, burst noise... are these types of noise related or are they all diffrent phenomenoms all together?
    Last edited: Jan 13, 2008
  2. jcsd
  3. Jan 14, 2008 #2
    Noise is as you said, unwanted disturbance in the signal. Noise is random and mathematically it is represented by a Gaussian distribution. There are various causes of noise, some of which you mentioned. The noise from each particular source is added up and the result is in the unit of mV. This can be +ve or -ve mV and is added to the calculation done without noise (used superposition here)

    Gray and Meyer is a standard textbook, but it's hard to read. If you are well versed with electronics and math then I suggest you buy the book. Another book is Mohammed Gausi (spelling?) or try Sedra and Smith (don't know if they have a chapter on noise)

    Or you can just google it: http://en.wikipedia.org/wiki/Electronic_noise
  4. Jan 14, 2008 #3


    User Avatar
    Science Advisor
    Gold Member

    Depends on what you mean by "noise". Unplebeian is right in that it is a "unwanted disturbance in the signal". However, much of the time the signal is corrupted due to interference (e.g. 50/60 Hz, mobile phones etc) which strictly speaking is not noise since it is not random.
    There are essentially 3 types of noise: The dominant source is usually thermal noise which is due to the temperature induced random fluctuations (i.e. it scales with temperature, which is why e.g. sensitive detectors are often cooled to low temperature). Shot noise is due to the discret nature of charge transport, i.e instead of a "stream" of electrons carrying a current you get something similar to hail, the "arrival time" of the charges is randomly distributed in time. Shot noise is a problem in e.g. GaAs transistors where it tends do dominate at low frequencies.
    Finally there is 1/f noise which gets its name because of the shape of the spectrum (it scales as 1/frequency). There are many sources of 1/f noise but it in general it is due to "telegraph noise" (noise sources descibed by the telegraph equation); essentially two-level fluctuators that jump back and forth.
    There is also quantum noise, but that can be ignored in almost all applications.
  5. Jan 16, 2008 #4
    Thank you both for your responses you have answered my question. I will continue to study this, just want to know the big picture.
  6. Jan 16, 2008 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    No true, for instance, in the case of switching noise.
  7. Jan 16, 2008 #6
    Yes, that is different, but he was not talking about common mode or differential noise in the case of switching power supplies.

    But it's good that you added those.
  8. Jan 16, 2008 #7


    User Avatar
    Science Advisor
    Gold Member

    Switching noise is just another name for the 1/f noise I mentioned above, it is a type of "intrinsic" noise. As the name imples the distribution is 1/f is frequency. In the time domain you get a number of "plateus" (i.e. the distribution is a set of sharp peaks, it is no where near gaussian) which corresponds to different configurations oftrapped charge (or whatever it is that is "jumping", it can also be e.g. flux).
    When doing sensitive measurements it is sometimes possible to see a single particle swiching between two states, in which case the distribution consists of two peaks.
    The switching can be slow; in some cases the time scale is of the order of minutes.
  9. Jan 16, 2008 #8
    True, true. What I meant earlier is to try to add to the information that will help the origial post instead of fighting/arguing between posts by two people.

    I'm sorry, but I've seen a lot of that lately on the forums (not geared towards you)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook