Analog to Digital Conversion Methods for Smart Sensing Systems

The new capabilities of smart sensing systems namely, adaptability, reconfiguration, lowenergy consumption and cost, between others, require a wisely selection of the methods that are use to perform analog to digital conversion. It is very important to optimize the trade-offs between, resolution, accuracy, conversion rate, and energy consumption, between others, and above all to adapt dynamically the conversion parameters for different signals characteristics and applications’ purposes. Establishing the best trade-offs are even more important when signals to be digitized have different signal-to-noise ratios (S/N) ratios, different requirements of measuring accuracy and acquisition rate, their characteristics are time-variant and above all if they are sharing the same digitalization device. Very low resolution or conversion rate of data acquisition (DAQs) systems are generally not compliant with measurement systems’ requirements since signal information is lost without any possible recovery procedure. Otherwise, if resolution or data acquisition rate are excessively high that means the sampling rate is much higher than its minimum value (Nyquist rate), the excessive amplitude and time resolutions provided by A/D conversion or frequency-to-digital conversion (FDC) does not improve measurements system’s performance. Moreover, the excessive resolution or data acquisition rate implies an increase of hardware and software complexity, data processing load and a higher implementation cost, without any benefits. So, for any A/D or FDC conversion method the best trade-off between different conversion characteristics must be established considering applications’ purposes. For example, in wireless sensing and actuating networks (WSAN) energy wastes are particularly important because a wrong choice of conversion method can affect deeply measurement system autonomy. Whenever possible, classical A/D conversion methods are being replaced by discrete A/D conversion methods that are supported by low cost microcontroller (C) (Microchip, 2010) connected to a few external resistive or capacitive components. This solution takes full advantage of Cs benefits, namely specific hardware and software capabilities and it provides a conversion rate that can be higher that several hundreds of kHz that is sufficient 24