I will have my Master's thesis about Bandpass Delta Sigma ADCs. I am currently in the learning phase, so I have a few questions :

1- I noticed that most of the design and simulation time is spent on Matlab. Does that mean I will not need to use Verilog AMS?

2- Since my focus is on analog/mixed signal design. How should I implement the digital filter?

3- Since the thesis is about Bandpass Delta Sigma ADCs. Will I need to learn something more than what is needed for designing normal Delta Sigmas. Ofcourse other than learning how Bandpass works  :)

4- Does the integrator actually consist of an Opamp or is an OTA preferred?

5- What is the block that I will probably spend most of my time designing?

Ok I have simple fundamental question. Why can't we do noise shaping unless we are oversampling?

I mean why not do noise shaping without oversampling. Why don't we use the same concept of noise shaping in Nyquist rate converters

I agree with that, but still, why not do noise shaping in nyquist rate converters to decrease the noise further?

Like Berti is saying noise shaping in nature requires oversampling.

To my knowledge noise shaping can only occur in the delta sigma structure where the quantization error is processed by the loop filter which in the case on the delta modulation is an high pass filter (integrator; delta modulation).

Therefore if the sampling frequency is much higher than than the signal frequency, noise is being greatly reduced or shaped in the band of interest (ie: signal band).

Ideally you can operate a delta sigma slightly above the Nyquist rate provided an ideal brick wall high pass loop filter which is pretty much impossible to built.

I would recommend the first chapter Richard Shreier book and the original paper on the delta modulation

[http://www.wiley.com/WileyCDA/WileyTitle/productCd-0471465852.html]

[H. Inose, Y. Yasuda, and J. Murakami, "A telemetering system by code modulation-Δ-Σ modulation." <i>IRE Transactions on Space Electronics and Telemetry, vol. 8, pp. 204-209, 1962]