Page 151 - The Master Handbook Of Acoustics
P. 151
126 CHAPTER SIX
pling pulses of Fig. 6-7B. These pulses, in effect, break down the analog
signal into a series of very brief samples having amplitudes equal to the
instantaneous value of the signal amplitudes, as shown in Fig. 6-7C.
This process is called digitization. With no loss of information, energy
between the sampling points is discarded. The sampling rate must be at
least twice the highest frequency of interest. If a sampling rate less than
this amount is used, spurious signals are generated.
Quantization
It is now necessary to convert the samples of Fig. 6-7C to discrete val-
ues that can be stored in a computer. This is done by a “sample-and-
hold” circuit so that the amplitude of each digitized pulse is converted
into discrete values suitable for computer storage. The sample-and-
hold circuit forces the amplitude of the sample to have a constant
value throughout the sample period. This process is essentially a step-
ping backward through a digitized signal, sample by sample, subtract-
ing from each sample some large proportion of the sample before it.
The resulting samples are thus mainly changes in the signal sample.
The closer the spacing of the digitized samples, the more accurately
the analog signal is represented. However, a restraining influence on
increasing the number of samples is that more computer memory is
required to store the data. The calculations required are intensive in
multiplication and accumulation operations.
Digital Filters
Digital filters can be made without benefit of inductors or capacitors. A
typical digital filter of the so-called FIR type is shown in Fig. 6-8. The
analog signal is applied to the input on the left. The analog-to-digital
(A/D) converter digitizes and quantisizes the analog signal. An oscilla-
tor (clock) determines the number of digitizing pulses per second and
controls all timing of the device. The type of filter is determined by the
program in the read-only memory.
Application of Digital Signal
Processing (DSP)
Digital signal processing has been successfully applied in various
1
ways, including :
