1
\$\begingroup\$

What happens to the high frequency part of a signal (in the region where the quantization noise is high) requantized with a delta sigma modulator, let's say with $$STF = 1$$ and $$NTF = (1-z^{-1})^N$$? I assume it is just lost into the shaped quantization noise, but does it affect also the stability of the modulator? In particular I know that MASH structures rely on the fact that the signal is still present inside the secondary modulator superposed to the added quantization noise. Does this "high frequency signal" (the quantization noise of the previous stage of the MASH) affect the stability of the modulator? The literature always assume the first order delta sigma is unconditionally stable, but also in presence of an input high frequency signal? My doubts arise from the fact that the high frequency part of the spectrum is used to stuff shaped noise, if this portion of the spectrum is already occupied by the signal itself does that mean that there is less "space" available and this could lead to stability problems?

\$\endgroup\$
0
\$\begingroup\$

There are no stability problems if the spectrum of a reasonable amplitude input signal extends into the 'noisy' region. The converter is stable in the presence of large voltage swings there anyway, a little extra additive input makes little difference.

With enough amplitude, it is possible to overload a sigma delta modulator, but this is just normal behaviour, any device has an overload level.

Given that the signal recovery filter after a sigma delta modulator removes, essentially completely, all the noisy region of the spectrum, it would be possible to overload the converter and not realise it, if looking at the filtered output signal alone.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.