TABLE OF CONTENTS The development of programmable digital logic and microprocessors and the availability of fast, reliable memory has made digital signal processing an attractive option in many applications that have hitherto depended on analogue techniques as well as opening up new possibilities that have not previously been viable.
Functions such as filtering, peak detection and frequency analysis can be performed numerically, often at lower financial cost as well as in terms of board area, component count and set-up time than for their analogue equivalents. Digital devices are essentially drift-free and calibration and alignment functions can be implemented in software very easily.
Digital signal processing solutions also offer the potential for in-the-field re-configuration or even self-adapting which can enhance maintenance and reduce the cost of ownership of products.
This chapter does not cover the detail of mathematics required to analyse and design digital filter and signal processing implementations or algorithms but is intended to convey the key points and implementation compromises that apply to digital signal processing using microcontrollers and DSP engines.
Analogue filters are usually linear time invariant systems and are characterized by the fact that they obey the principle of superposition, that is the output resulting from applying the sum of a set of independent signals to the input of a filter is the same as summing the outputs obtained for all the signals applied independently. This being the case, we can characterize a filter by its step response the output response to...
