Page 1158 - The Mechatronics Handbook

FIGURE. 44.1

Analog signals for data acquisition can be grouped into two basic classes: random and deterministic.
Data acquisition devices can both acquire and generate these types of signals. Random signals never repeat
and have a flat frequency spectrum. Microphone static is an example of a random signal. A deterministic
signal, unlike random signals, can be represented by a sum of sinusoids. Deterministic signals can be
subdivided into periodic and transient signals. Periodic signals constantly repeat the same shape at regular
intervals over time, while transient signals start and end at a constant level and do not occur at regular
intervals. Transient signals are nonperiodic events that represent a finite-length reaction to some stimulus.
Digital input and output are commonly incorporated into data acquisition hardware for sensing
contacts, controlling relays and lights, and testing digital devices. The most commonly used digital levels
are TTL and TTL-compatible CMOS. These are both very common 5-V standards for digital hardware.
Digital transfer rates to and from the data acquisition hardware vary from unstrobed to high speed.
Unstrobed digital input and output involves setting digital lines and monitoring states by software
command. This form of digital input and output is also known as static or immediate digital I/O. The
maximum speed of an unstrobed I/O is highly dependent on the computer hardware, the operating
system, and the application program. Pattern digital I/O refers to inputs and outputs of digital patterns
under the control of a clock signal. The speed at which the data can be sent or received depends on the
amount of data, the characteristics of the data acquisition hardware, and the computer speed.
The final type of I/O on computer-based data acquisition hardware is counter/timer I/O. Counter/
timers are capable of measuring or producing very time-critical digital pulses. These pulses, like the
digital input and output, are generally TTL or TTL-compatible CMOS. These components are used for
measuring or producing a number of time-critical signals including event counting, pulse train genera-
tion, frequency-shift keying, and monitoring quadrature encoders. The two main characteristics of a
counter/timer are the counter size and maximum source frequency. The counter size is generally repre-
sented in bits and determines how high a counter can count. For instance, a 32-bit counter can count
32
2 - 1 = 4,294,967,295 events before it returns the count value back to zero. The maximum source
frequency represents the speed of the fastest signal the counter can count. An 80-MHz counter can count
events that are as fast as 12.5 ns apart. An “event” is actually the rising or falling edge of a digital signal.
No real situation will ever have perfect signals or be completely free of noise. Signal conditioning is a
method to remove, as much as possible, unwanted components of a digital or analog signal. A real analog
signal usually comprises both deterministic and random signals, and a digital signal is not going to be
perfectly square. Measurement hardware, particularly for high-frequency analog signals, is usually
equipped with an antialiasing filter. This is a low-pass filter that blocks frequencies above the desired
frequency range and increases the accuracy of the measurements. Digital and counter/timer lines are also
commonly fitted with filters that remove spikes from the signal that could otherwise be mistakenly

©2002 CRC Press LLC