An ARMA-Model-Based NTF Estimation on Continuous-Time ≪inline-Formula> ≪tex-Math Notation="latex">$\delta\sigma$</tex-Math></inline-formula> Modulators

Chixiao Chen,Zemin Feng,Jun Xu,Fan Ye,Junyan Ren
DOI: https://doi.org/10.1109/tcsii.2015.2415272
2015-01-01
IEEE Transactions on Circuits & Systems II Express Briefs
Abstract:In this brief, a noise transfer function (NTF) estimation algorithm is developed for continuous-time delta-sigma modulators. Different from previous methods, the proposed estimation method features an autoregressive-and-moving-average-model-based NTF approximation and low-cost computational resource. The NTF's zeros are derived by a spectral estimation on the quantization noise under the special situation, whereas the poles are obtained by a Feintuch-LMS adaptive filter with out-of-band tone injection. The proposed method could also track imprecise compensation effects on excess loop delay and finite gain and bandwidth. The parameter number of the method is only 2N+1 for an Nth-order modulator. The simulation result shows that the proposed estimation method fits the modulator after compensation with less than 1.7 × 10 -6 V 2 mean square error with ±0.6-Vpp swing in a third-order modulator.
What problem does this paper attempt to address?