Measuring signal to noise ratio

[vreg]

Hi,

Can anyone explain how measurements of signal to noise ratio (SNR) made?

I read that a signal source is used to generate a signal and then at the output of the receiver, a voltmeter is used to measure the received signal level.

But won't this method include noise as well, i.e. isn't it a signal plus noise (S+N) measurement, rather than SNR (S/N) measurement?

 

[KlausST]

Hi,

Nowadays SNR often is calculated by using FFT.

In the simplest way you see the signal frequency amplitude, and you see all the noise amplitudes. Calculate the RMS value of all noise and calculate the relation to your signal RMS amplitude. Finished.

If you expect harmonic distortion you should exclude the overtone frequencies when you calculate the noise amplitude.

Some use additional weighting filters. Especially for audio signals.

DC is considered as no frequency so it is not included in noise amplitude.

You may also work without real "signal", but consider it to be full scale RMS.

For SINAD you need a real signal frequency and you need to include overtone frequencies in calculations.



Klaus

 

[vreg]

Hi,

So the method just measuring the voltage at the receiver isn't valid right?

I had the following questions regarding the method you described:
To calculate SNR using FFT you will first have to convert the signal to the digital domain- so the final SNR that you get is after adding the noise of the entire signal path between the antenna and the digital base-band to the noise introduced by the channel, i.e. (Actual signal strength/(Channel Noise + Receiver Noise))?
Also, what if the noise also occurs at the frequency of the signal, how do you account for that?

If you expect harmonic distortion you should exclude the overtone frequencies when you calculate the noise amplitude.

Can you elucidate this please. I know what are harmonics, but what do you mean by "expect harmonic distortion"?

Thanks

 

[KlausST]

Hi,

I just recognized we are in the RF area. My answer is more for lower frequencies. Sorry for that.
I´ve very limited experience with RF.

I can think of a narrow bandpass to get the "signal of interest"
and a band stop filter to get the noise
and set both RMS values in relation.

Klaus

 

[chuckey]

Signal to noise ratio is specified at a certain signal level. Often for communication receivers it at 1 microvolt. So using a wattmeter on the output and with the specified input level, set the output to some convenient level, remove signal, see how much noise there is. This gives you S + N/N, if you want S/N you will have to calculate it, because you can't measure the signal without the noise.
Frank

 

[D.A.(Tony)Stewart]

Some receiver manufacturers will specify S+N/N . In practice the difference is not large, but the S+N/N ratio is more correct.

The input signal must be the Loaded signal ( match to input) usually stated in dBuV (phone lines) or dBm ( radios) depending on application and not the no load voltage (Vemf or EMF) which is double.

in radios where there may be distortion, then SINAD is preferred.

SINAD = 10Log ( SND / ND )
where:
SND = combined Signal + Noise + Distortion power level
ND = combined Noise + Distortion power level

 

[vreg]

What do you mean by "The input signal must be the Loaded signal "?

 

[vfone]

For calculating the required SNR of a receiver (at the input of the detector), you have to know the sensitivity of the receiver, the receiver bandwidth, and the receiver noise figure.

SNR[dB] = 174 + RX_Sensitivity[dBm] – 10*LOG(RX_BW[Hz]) – RX_Noise-Figure[dB]

 

[vreg]

Hi,

Is the RX_Sensitivity related to gain of the antenna in any way? And what is the intuition behind including the RX_BW in the formula?

Also where did 174 come from?

Thanks