Passive Loop Filter Assistance for CTSDMs

Abstract

This paper presents a power reduction technique for continuous time sigma delta modulators (CTSDM). The approach consists of two elements. First, a passive low pass filter is added in front of the modulator's loop filter to reduce the high frequency components in the loop. As a result, the slew rate requirements of the opamps can be greatly reduced which allows a significant power saving. Unfortunately, the insertion of this low pass filter also changes the modulator's loop gain, and hence affects the NTF and STF (Noise-and Signal Transfer Functions), in an undesired way. Therefore, the second proposed element consists of inserting a compensation branch which is such that the original loop gain, NTF and STF are restored. Thanks to this, our power saving technique is completely transparent on the system level such that all established techniques and toolboxes for CTSDM design can still be used. The technique is especially suited for one-bit CTSDMs where the amount of high-frequency components in the loop is excessive. To showcase the technique, an SOSDM (which is a dedicated type of one-bit CTSDM) was implemented in a 65nm CMOS process. It achieves a peak SNDR of 63dB over a 20MHz bandwidth at a power consumption of 1.7mW while occupying a very small chip area of only 0.009 mm 2 .

Low-Pass Filtering SC-DAC for Reduced Jitter and Slewing Requirements on CTSDMs

“…We will denote this DAC structure as a low-pass filtering switched capacitor DAC (LPSC-DAC). However, similar as in other filtering techniques [3], the loop's noise-and signal transfer functions are again affected and we explain how this can be solved by a compensation scheme similar to the one of [3].…”

“…As will become clear later on in section IV, the introduced compensating zero can be a lot lower than the clock frequency, which would significantly change the filtering characteristic of the STF. For full compensation, it is therefore necessary to add a low-pass filter in front of the loop, as was also done in [3] for a system with an NRZ-DAC. Fig.…”

“…First variants of the analog filtering approach were proposed in [3] and [11]- [13]. It consists of adding a passive low-pass filter in front of the first OpAmp in the loop filter.…”

“…Unfortunately, adding this filter to the loop puts constraints on the modulator's noise transfer function (NTF) such that the noise shaping is affected. Solutions for this problem have been proposed in [3] and [12]. In the former, an extra zero is added to the loop filter after the first OpAmp to cancel the pole introduced by the low-pass filter, leading to the system of Fig.…”

“…1. As a result, the technique of [3] is completely transparent on the system level, leaving the loop filter, noise transfer function (NTF) and signal transfer function (STF) unchanged. Unfortunately this approach does not improve errors introduced at the DAC, like jitter or ISI.…”