Look-ahead decision-feedback ΣΔ modulation

We present a novel solution to the problem of instability in multiloop ΣΔ modulators (nonlinear oversampled one-bit A/D and D/A converters). ΣΔ modulators are attractive because their SNR is proportional to their oversampling rate which is physically tied to the speed of the VLSI. The problem of instability in traditional higher order (multiloop) ΣΔ modulators is corrected by a new class of self-stabilizing ΣΔ modulators called look-ahead decision-feedback (LADF) ΣΔ modulators. LADF ΣΔ modulators select the output at each time-step to create the optimally stable set of integrator states at the following time-step. Simulation results for up to third order (triple loop) modulators are included