Noise is constant or changing?

Discussion in 'General Electronics Chat' started by Jas9, Feb 7, 2011.

  1. Jas9

    Thread Starter Member

    Sep 30, 2010
    Hello friends,
    i want to know whether a noise or offset produced by a sensor for measuring for example voltage will remain constant or will keep on changing?
    For example: I have a sensor which measures the voltage of a battery continuously. If the real voltage is 12V and because of noise in the sensor it indicates as 12.2 V. So now we see and offset of +0.2V. So for the next measurement if it has to measure 12.5V then what it will it indicate 12.7V? or !2.something else. Because according to me i think the offset of sensor (in this case +0.2V) will remain constant over a period of time and maybe due to some breakdown in the sensor this offset will get increased.
    So please let me know the real fact and answer for the above.
  2. timrobbins

    Senior Member

    Aug 29, 2009
    What sensor are you using, and in what measurement circuit? Maybe you are confused about what 'noise' is, compared to an 'offset' - but that all depends on what your circuit and sensor is.

    Ciao, Tim
  3. Jas9

    Thread Starter Member

    Sep 30, 2010
    well to understand the concept i placed that example...but how it is in reality? i mean if for example the values to be measured are [12 12.5 11.94 11.5 13.5 14.4] but if a sensor has some noise then will then noise effect will be same on all measured reading that is if the noise level is +0.2 then will it be same through out the measurement or +0.2 will be changed for each reading?
  4. tom66

    Senior Member

    May 9, 2009
    There are three major sources of error.

    1st. Offset, as you mention. 12.2 V instead of 12V. But also 0.2V instead of 0V.

    2nd. Accuracy - which works alongside error as a term in y=mx + c. Basically, accuracy is given as a percentage tolerance or parts per million tolerance. So a voltage sensor with an error of ±5% may read anywhere from 11.4V to 12.6V, before any offset is added. Crucially, accuracy is not a simple offset term. It would be off by zero at 0V, and by 5V at 100V, if it were ±5%.

    3rd. Temperature and other environmental conditions may cause the above to drift making calibration less accurate.
  5. Wendy


    Mar 24, 2008
    All of what Tom has said is true, noise is actually a fourth variable. Some electronics have a tendency to generate is internally. If it is true noise, it is random, it is the "hiss" you hear over a radio (which is a reflection of signal to noise ratio). Expensive parts and good designs minimize it. Sometimes it is wanted, to make measurements such as S/N ratios. Zeners can generate low levels of it (but not much). Zeners tend to make good power supplies, or power supply voltage references, so their noise must be supressed with some kind of filter arrangement.
  6. Jas9

    Thread Starter Member

    Sep 30, 2010
    thanks to you all for a great advice
  7. Papabravo


    Feb 24, 2006
    I like to think of noise as an imprecise phenomena with statistical properties like mean and variance. Rhetorically one could ask about the difference between a "constant" K, and a random variable with a mean of "K" and a variance that approaches 0.
  8. someonesdad

    Senior Member

    Jul 7, 2009
    There are many different physical sources and types of noise and the correct answer to your question, in general, is you can't know without making measurements and characterizing the noise. Even if you characterize it and decide on some statistical moments, you don't know if those characteristics are constant in time. Measurement is the only way to know.

    If all you have is e.g. a digital multimeter, about the best you can do is to make many measurements and calculate the mean and standard deviation and use those to characterize the variability of the signal. I'm a big fan in plotting the data -- both as a time series and on normal probability paper. Those two plots can give you quite a bit of information about the characteristics of the "noise".