Basic Pulse Circuits

Part 1: Essentials for Processing Outside Signals 9

- Wolfgang Matthes

In part one of a three-part series, Wolfgang writes how basic pulse circuits help digital circuits, such as embedded boards with ARM processors, deal with pulse trains or bursts of pulses from the outside. He delves into design tasks needed to capture, detect, filter, and synchronize pulses, essential subjects many textbooks don't cover.

Many digital systems must process signals from outside, that is, from sensors, level translators, and the like. Most of them are essentially pulse trains or bursts of pulses. They are to be processed by microcontrollers, programmable logic, or application-specific circuits. Various problems may arise. Some pulses may be too narrow, some too wide, pulse trains may be affected by glitches, and so on. Typical design tasks concern capturing pulses, detecting edges, filtering out or detecting glitches, and synchronizing the pulses with internal clocks.

These are minor but intricate problems most textbooks do not cover. Solving such problems requires some familiarity with the appropriate components and some inclination to deal with the utmost details. Purely highlevel approaches will not be useful here. This is also true for the cookbook solutions of the past and the textbook theory of sequential circuits [1]. In this first of three parts, we will introduce the latches and flip-flops as the most basic components and discuss fundamental design problems.

Figure 1 shows a signal line feeding a pulse train to a digital IC. Typically, the pulses are asynchronous signals, not related to the internal timing, especially to the clocks. They come from sensors, mechanical contacts, industrial application environments, and the like. The signals may be subject to coupled-in noise and crosstalk. Other spurious, irregular pulses may result from the very nature of the signal source.

Think, for example, of bouncing contacts.