“…After considering circuit stability and spectral tone rejection, a 2-1 cascaded architecture was adopted instead of the single-loop architecture. This architecture has advantages compared to the 1-1-1 cascaded topology, as follows: first, it is unconditionally stable if all of its stages are unconditionally stable, as are the case for the second and first order stages; second, the use of second order noise shaping in the first stage of the cascade avoids the potential presence of discrete noise tones in the overall cascade output, which is desirable for high performance audio applications; third, it is less sensitive to component mismatch [3,4], especially when the feature size is scaled down to 65 nm, where the circuit faces ever greater challenges of variability and reliability.…”