9.2 Analog-to-Digital Conversion: General Issues

Although Comparator_A produces digital values from some types of sensor, many applications call for a "real" analog-to-digital converter (ADC). I explain some general features of ADCs here and go on to describe features of the specific types of ADC provided in the MSP430 in subsequent sections.

The purpose of an ADC is to convert an analog input, which I take to be a voltage V, into a binary value that the digital processor can handle. The input V(t) is a continuous function, meaning that V can take any value within a permitted range and can change in any way as a function of time t. In contrast, the output V[ n] is a sequence of binary values. Each has a fixed number of bits and can therefore represent only a finite number of values. Typically the input is sampled regularly at intervals of T _s, so the continuous nature of time has also been lost. Thus the process of conversion damages both V and t and we examine the effects next.

Mixed-signal systems present serious challenges and it is no surprise that whole books are devoted to them. I find A Baker's Dozen: Real Analog Solutions for Digital Designers [39] to be the most accessible and the author regularly covers mixed-signal issues in her column in Electronic Design News. There are also good books on data conversion from Analog Devices [43, 44] and books...