Systematic Design Exploration of Delta-Sigma ADCs

Abstract

Abstract-An algorithm for architecture-level exploration of the 16 A/D converter (ADC) design space is presented. Starting from the desired specification, the algorithm finds an optimal solution by exhaustively exploring both single-loop and cascaded architectures, with a single-bit or multibit quantizer, for a range of oversampling ratios. A fast filter-level step evaluates the performance of all loop-filter topologies and passes the accepted solutions to the architecture-level optimization step which maps the filters on feasible architectures and evaluates their performance. The power consumption of each accepted architecture is estimated and the best top-ten solutions in terms of the ratio of peak signal-to-noise+distortion ratio versus power consumption are further optimized for yield. Experimental results for two different design targets are presented. They show that previously published solutions are among the best architectures for a given target but that better solutions can be designed as well.Index Terms-Analog-to-digital converter (ADC), analog-to-digital, computer-aided design (CAD), delta-sigma, design automation.

Hierarchical statistical analysis of complex analog and mixed-signal systems

“…In Kuo et al (1999), Bajdechi et al (2004), Wang and Kuo (2006) and Tang and Doboli (2006), the assumption of independent Gaussian distribution of the signal path coefficient in the Delta-Sigma modulator may not be realistic, since the coefficients are implemented in different circuits that are subject to variations of common process parameters. Also, the circuit-level non-idealities should be considered for more accuracy since the real SNR variation is dependent on them.…”

“…Besides, the variation of design variables (such as resistors and capacitors) are not related to process parameter variations. Similar work has been also reported on statistical analysis of Delta-Sigma modulators (Bajdechi, Gielen, & Huijsing, 2004;Kuo, Chen, & Chen, 1999;Tang, 2007b;Tang & Doboli, 2006;Wang & Kuo, 2006;Webb & Tang, 2008). An ideal behavioural model, such as the one shown in Figure 1 for a fourth-order continuous-time Delta-Sigma modulator, is used.…”

“…In this case, SNR also follows Gaussian distribution due to the linear equation. Kuo et al (1999), Bajdechi et al (2004), Wang and Kuo (2006) and Tang and Doboli (2006) assume random Gaussian or uniform distributions for all signal path coefficients and directly apply the Monte-Carlo method for statistical analysis. However, the above approaches do not relate performance parameter variations to fundamental process parameter variations plus many circuit-level nonidealities are not considered; therefore, the statistical analysis in those approaches may not be realistic.…”

Designing globally optimal delta–sigma modulator topologies via signomial programming

“…Simulations combined with genetic algorithms are then employed to select the right topology. Bajdechi et al [12] use behavioral simulation to evaluate the power consumption of possible candidates: several candidate solutions are subsequently passed to Monte Carlo analysis to find the best topology that gives the highest yield. However, the main limitations of these approaches are (1) Only a finite number of topologies are considered.…”

Design of interpolative sigma delta modulators via a semi-infinite programming approach

“…The popularity of SDMs is mainly due to their simple, inexpensive and robust circuit implementations, as well as achieving very high SNR because of their ability to perform noise shaping [4]. SDMs are typically designed using Butterworth filter design rules [5], and optimal designs have been performed via HSpice [6], Matlab [7] and Fortran [8]. Although these designs have considered many practical issues, the solutions obtained are not the global optimal one.…”