Biomedical Engineering Reference
In-Depth Information
1.5
1
0.5
0
T
-0.5
-1
-1.5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
FIGURE 6.5
:
Digital signal sampled at a rate of 1
/
T
, where
T
is the sampling interval
as a sequence of numbers
x
(
n
) representing sampled values. Figure 6.5 shows a sampled
sinusoid with a sampling interval of 1/10 the period of the sinusoid.
v
(
t
)
=
sin(
t
)
T
=
2
π/
10
6.3 QUANTIZATION
Now, let us consider the quantization operation since the magnitude of each data sample
must be expressed by some fixed number of digits, since only a fixed set of levels is
available for approximating the infinite number of levels in continuous data. The accuracy
of the approximating process is a function of the number of available levels. Since most
digitizers produce binary outputs compatible with computers, the number of levels may
be described by the number of binary digits (bits) produced.
6.3.1 Resolution or Quantization Step
In digital systems, sampled values must be represented in some binary number code that
is stored in finite-length register. A binary sequence of length
n
can represent at most
2
n
different numbers: that is, a 4-bit register can represent 2
4
or 16 numbers from 0 to
15 inclusively. If the amplitude of the analog signal is to be between
−
v
and
+
v
volts,
Search WWH ::
Custom Search