Signal to noise ratio

[c.fish]

Hi to everyone,

Is there anyone that could explain me how to practically calculate the SNR of a vibrational signal?

I know the mathematical expression, that is:

SNR=signal power/noise power

but i do not know how to apply it.

[johnsold]

Measuring and calculating SNR depends on the details of your system.

Is the desired signal a simple signal with a single frequency? Or does the frequency vary within the measurement interval? Does the signal have a complicated waveform or multiple frequency components?

Are you applying an external excitation signal or does the system generate the signal internally? Do you know if the noise is mostly random or are there spurious components at frequencoes other than the desired signal frequencies? Are there resonances in the system? Do the resonances occur at frequencies of the desired signals?

The basic measurement/calculation issue is how to separate the signal power from the noise power. That is why you need to understand the characteristics of the system.

Lynn

[c.fish]

Hi johnsold, and thank you for your aswer.

My signal comes from an accelerometer placed on the engine block of an internal combustion engine; so it has a very complicated waveform.

I think that the noise is mostly randon, namely it is produced by everything that do not belong to the gas evolution in the combustion chamber.

Do you have any idea?