Noise and SNR.

[Zener_Diode]

hey

SNR=20*log(A-signal/A -noise)

what is A-noise? how can I measure the noise amplitude?
what statistics calculation should I do to measure noise amplitude??

thanks ahead.

 

[Sceadwian]

Try Wikipedia's entry on Signal-to-noise ratio.
https://en.wikipedia.org/wiki/Signal-to-noise_ratio

It's best to look something up before you ask a question about it, especially if you're in school.

 

[Zener_Diode]

Try Wikipedia's entry on Signal-to-noise ratio.
https://en.wikipedia.org/wiki/Signal-to-noise_ratio

It's best to look something up before you ask a question about it, especially if you're in school.

I read it, no information related to my question.

 

[Sceadwian]

Strange the second part of the wikipedia entry I looked up answers every question you asked....

The following is a quote from the wikipedia entry.

Technical sense

Signal-to-noise ratio is an engineering term for the power ratio between a signal (meaningful information) and the background noise:
**broken link removed** where P is average power and A is RMS amplitude. Both signal and noise power (or amplitude) must be measured at the same or equivalent points in a system, and within the same system bandwidth.
Because many signals have a very wide dynamic range, SNRs are usually expressed in terms of the logarithmic decibel scale. In decibels, the SNR is, by definition, 10 times the logarithm of the power ratio. If the signal and the noise is measured across the same impedance then the SNR can be obtained by calculating 20 times the base-10 logarithm of the amplitude ratio:
**broken link removed**

 

[Zener_Diode]

Where is the answer to:
How to measure the noise amplitude???
Noise is a random numbers, they don't have an amplitude. I should to calculate the amplide somehow.
how can I do it????

 

[Hero999]

It is a random number within set bounds.

The noise signal is a random number between -1 and 1 times the noise amplitude.

The noise signal is simply added on to the useful signal and the ratio between the two is the signal to noise ratio.

 

[Sceadwian]

Read the entry very slowly.
A = RMS amplitude.

The same method is used to measure both signal and noise, you just have to use the same number of samples. Noise is measured simply by applying NO signal.

The simplest explanation I can give is to measure the signal to noise ratio of a human voice in any given room, you take two samples, the room by itself, and then the room with the signal in it. You RMS the samples of both of those readings and apply the algorithm and you have your SNR for those given samples.

Please refer to...
https://en.wikipedia.org/wiki/Root_mean_square

 

[Zener_Diode]

I just what?

what I did was:
shorted the input, increse the gain till I could measure the noise and output in to excel file.
Now I have a list of numers that represents me a noise at gain G1, if I divided by G1 and multiply it by the real gain I have, I should get the noise at my gain.
I want to caculate SNR. What is my A-noise????

 

[Sceadwian]

I'm sorry Zener, we seem to be having a very basic communication issue here.

As I stated above and is clearly stated in the wikipedia entry. A is the RMS amplitude of the measured samples. In order to determine the SNR ratio you need to determine the RMS value of all of those samples you took. That's why I pointed you to the RMS entry on wikipedia as it contains all the equations required to determine an RMS value from a given set of readings. Those numbers might appear random, but if you apply the RMS algorithm to it you're going to get a non zero value EVERY time with a real world instrument. That then becomes your A-noise reading. You then repeat your measurements with the signal to be compared against and run that through the RMS algorithm. Do not change the gain or it completly invalidates your results as the A-noise and the A-signal readings have to be taken under identical circumstance, the only difference allowed is the application of the signal to be measured against. If when the gain is set to properly measure your normal input signal you can not measure any value when the input is shorted then your SNR ratio is bellow the quantization error of your signal, in which case the SNR is calculatable from the bit depth of your generated signal directly using the equations in the "Fixed Point" topic in the wikipedia entry for SNR.

 

[eng1]

Noise is a random numbers, they don't have an amplitude. I should to calculate the amplide somehow.
how can I do it????

Noise is described by the power spectral density, that is a function of the frequency. If you want to calculate the rms voltage in a given interval [f1, f2], integrate the spectral density and take the square root.
For example, thermal noise in a resistor (simplified model) has a flat spectral density, that equals 4*k*T*R k=1.38e-23 (Boltzman constant), T=temperature in Kelvin, R=resistance. The rms voltage is sqrt[4kTR(f2-f1)].
Noise calculations in complex circuits are difficult, though.

 

[Sceadwian]

He has Excel entries for his data eng1. Simply reading the RMS entry on wikipedia and applying that to the N set he has and an equivalent signal sample will give him his answer.

 

[eng1]

I realized that atfer finishing my reply, Sceadwian!

 

[3iMaJ]

Noise amplitude is the standard deviation. Noise power is the variance.

 

[Sceadwian]

Am I speaking in tongues or something? A in that equation is RMS voltage that's for both noise and signal amplitude. Here's a quick excel equation to calculate the RMS value of a range of entries.

=SQRT(SUMSQ(A1:A17)/COUNTA(A1:A17))

That field in Excel will return the RMS value of entries A1 through A17

To calculation signal to noise ration you need two samples.
The measured signal, and the measured noise, and they must be measured under the same circumstances or the value means nothing. The sample size will dramatically effect the results.

 

[audioguru]

An "A-contour filter" is frequently used in noise measurements because it closely resembles the frequency response of our hearing at very low levels.

I think it is cheating if you measure the noise of a circuit with its input shorted. The input should be terminated with the normal source impedance.