No abstract

A ∆Σ ADC using an LSB-first quantizer (LSBFQ) is proposed. LSBFQ are energy-efficient ADCs for processing signals with low activity, and have been proposed as standalone quantizers for sensor and biomedical applications. Since the quantizers in highly oversampled multibit ∆Σ ADC process signals with low average activity, the LSBFQ is an ideal quantizer solution. In order to avoid clocking the LSBFQ at a rate much faster than the rest of the ∆Σ ADC, it is proposed that the quantizer be provided a fixed number of comparison cycles, then be interrupted regardless of whether the conversion has fully completed. This is acceptable because for high oversampling ratios (OSR), the average code change is small and an N -bit conversion can usually be completed in fewer than N comparison cycles. In the rare cases that the quantizer is interrupted early, it injects slightly more quantization noise into the loop filter, which is filtered and shaped with little impact on signal-to-noise and distortion ratio (SNDR). Simulation results demonstrate that for high OSR, an LSBFQ achieves higher resolution and lower capacitor switching energy than a conventional SAR ADC using the same number of comparator bitcycles.

Prior work used least-significant bit first quantization (LSBFQ) to conserve switching energy and comparator bitcycles, but is limited to low activity signals. Furthermore, LSBFQ results in a large bitcycle range in the quantizer. A novel selectable starting bit quantizer (SSBQ) is proposed which starts quantization with neither the MSB nor the LSB, but an intermediate bit chosen for target applications. It is shown that the proposed algorithm reduces bitcycle range in the quantizer compared to LSBFQ, and provides design flexibility for various activity signals. Furthermore, the proposed solution encompasses LSBFQ since it is a specific case of the proposed architecture. For target applications, the proposed solution will save bitcycles in an A/D conversion, as well as switching energy, over the LSBFQ and the merged capacitor switching (MCS) SAR, the most energyefficient traditional MSB-first SAR.