Tweet
It is generally known that adding white noise signal gives 1/2 bit resolution each time you double the number of samples.
Another way to state this is that n samples reduce the noise N by the square root of the number of samples
But who knew that a simple low-pass IIR filter y_t = (n - 1) y_{t-1} + x_t improves that by a factor
?
\frac{N}{\sqrt{2n-1}}
From this article: https://www.electronicdesign.com/tec...duce-adc-noise
But it can get even better, who knew that adding triangle noise gives you +1 bit improvement each time you double the number of samples?
})
From this application note by ST https://www.st.com/resource/en/appli...lectronics.pdf
Another way to state this is that n samples reduce the noise N by the square root of the number of samples
But who knew that a simple low-pass IIR filter y_t = (n - 1) y_{t-1} + x_t improves that by a factor
\frac{N}{\sqrt{2n-1}}
From this article: https://www.electronicdesign.com/tec...duce-adc-noise
But it can get even better, who knew that adding triangle noise gives you +1 bit improvement each time you double the number of samples?
From this application note by ST https://www.st.com/resource/en/appli...lectronics.pdf
Comment