Introduction - Introduction to Digital Signal Processing and Filter Design - page 28

TABLE 1.3 ADCs Currently Available

Maximum
Frequency in
Input Signal

Sampling Rate
(samples per second)

Resolution
(bits)

Power

96,000 24 48 kHz 90 mW

96,000 18 48 kHz 60 mW

96,000 16 48 kHz 40 mW

65,000,000 14 500 MHz 0.6 W

400,000,000 8 1 GHz 3 W

frequency of the ADC and DAC that are available. As the frequency is increased,
the wordlength of these devices decreases and therefore the accuracy and dynamic
range of the input and output data decrease.
For example, data on a few ADCs currently available are given in Table 1.3 [3].
Hence digital signal processing is restricted to approximately one megahertz,
and analog signal processors are necessary for processing signals above that
frequency, for example, processing of radar signals. In such applications, analog
signal processing is a more attractive and viable choice, and currently a lot of
research is being directed toward what is known as mixed-signal processing. Note
in Table 1.3, that as the resolution (wordlength) for a given signal decreases, the
power consumption also decreases, but that is just the power consumed by the
ADCs; the power increases as the sampling frequency increases, even when
the resolution is decreased. The digital signal processor itself consumes a lot
more power, and hence additional power management circuits are added to the
whole system. In contrast, the analog signal processors consume less power.
The LC filters consume almost negligible power and can operate at frequencies
in the megahertz range. The active- RC filters and switched-capacitor filters are
restricted to the audiofrequency range, but they consume more power than do
the LC filters. It is expected that mixed-signal processing carried out on a single
system or a single chip will boost the maximum frequency of the signal that can
be processed, by a significant range, beyond what is possible with a strictly digital
signal processing system. Therefore we will see more and more applications of
DSP with increasing frequencies because the advantages of DSP outweigh the
disadvantages in analog signal processing.

1.6 SUMMARY

In this introductory chapter, we defined the discrete-time signal and gave a few
examples of these signals, along with some simple operations that can be applied
with them. In particular, we pointed out the difference between a sinusoidal
signal, which is a continuous-time signal, and a discrete-time signal. We dis-
cussed the basic procedure followed to sample and quantize an analog signal