Multi‐Stage and Multi‐Quantizer Delta‐Sigma Modulators
Richard Schreier, Shanthi Pavan, Gabor C. Temes
Abstract
Richard Schreier, Shanthi Pavan, Gabor C. Temes
Abstract
This chapter discusses delta-sigma modulators that combine noise cancellation with noise-shaping and examines how the signal-to-quantization noise ratio (SQNR) can be increased by increasing the oversampling ratio (OSR), oversampling ratio, or the quantizer resolution, or the order or aggressiveness of the loop filter. A different strategy is to cancel the quantization noise in addition to noise filtering using a multi-stage structure for the modulator. The quantized outputs can then be combined in a way that reduces the overall noise power. In high-order single-stage modulators, the imperfect matching of the passive loop filter elements and the finite gain of the active ones will change the coefficients of the NTF and STF. As is usual for delta-sigma modulators, accurate behavioral simulation is the most reliable technique for predicting the effects of all non-idealities on the SQNR of a MASH modulator. Computer simulations indicate that the performance of the modulator remains minimally affected when the opamp dc gains are as low as 30 dB and the element errors are as large as 5%.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This chapter discusses delta-sigma modulators that combine noise cancellation with noise-shaping and examines how the signal-to-quantization noise ratio (SQNR) can be increased by increasing the oversampling ratio (OSR), oversampling ratio, or the quantizer resolution, or the order or aggressiveness of the loop filter. A different strategy is to cancel the quantization noise in addition to noise filtering using a multi-stage structure for the modulator. The quantized outputs can then be combined in a way that reduces the overall noise power. In high-order single-stage modulators, the imperfect matching of the passive loop filter elements and the finite gain of the active ones will change the coefficients of the NTF and STF. As is usual for delta-sigma modulators, accurate behavioral simulation is the most reliable technique for predicting the effects of all non-idealities on the SQNR of a MASH modulator. Computer simulations indicate that the performance of the modulator remains minimally affected when the opamp dc gains are as low as 30 dB and the element errors are as large as 5%.
Key concepts: Oversampling, Delta-sigma modulation, Noise shaping, Quantization (signal processing), Noise (video), Electronic engineering, Control theory (sociology), Computer science