Digitally enhanced digital-to-analogue converters

Abstract

With contemporary electronic systems processing information primarily in the digital domain, digital-to-analogue converters (DACs) play an essential role in interfacing to the physical world which often require a fine-grained range of signal levels being applied to the physical domain in question. DACs, therefore, are an integral part of driving actuators such as antennas for radio transmission, audio speakers for generation of sound or implanted electrodes for therapeutic stimulation of nervous tissue. DACs, on the other hand, are also used internally in electronic systems such as parts of analogue to digital converters (ADCs) or in built-in test systems. DACs are used with a wide variety of resolutions - from a single bit to tens of bits - over a wide variety of conversion speeds - from kHz ranges to GHz ranges. While low-resolution DACs are fairly straightforward to design by relying on component matching, the performance of medium and high-resolution DACs is limited by component variation. In this chapter, we use digital techniques - specifically dynamic element matching (DEM) - to improve DAC performance beyond the limitations imposed by component matching. The authors begin by giving a brief overview of common DAC encoding schemes, error mechanisms and error metrics; they also introduce a 32-element DAC structure that will be used for illustration purposes. They discuss techniques for DAC error mitigation and linearisation; in particular, the use of DEM where they use the 32 unit element DAC structure to illustrate the operation of various DEM techniques. They look at a case study of a harmonic-cancelling DAC (HC-DAC) used for sine-wave synthesis and do a detailed analysis of mismatch errors and the effects of DEM; they further present measurements from an HC-DAC with partial DEM.