Hardware Reference
In-Depth Information
Many A/D conversion algorithms have been introduced in the past. These algorithms can
be divided into four categories.
1. Parallel (Flash) A/D converter
2. Slope and double-slope A/D converter
3. Sigma-delta A/D converter
4. Successive-approximation A/D converter
P
ARALLEL
(F
LASH
) A/D C
ONVERTERS
In this type of A/D converter, 2
n
comparators are used. One of the inputs to each comparator
is the input voltage to be converted; the other input corresponds to the voltage that represents
one of the 2
n
combinations of
n
-bit values. The comparator output will be high whenever the
analog input (to be converted) is higher than the voltage that represents one of the 2
n
combina-
tions of the
n
-bit value. The largest
n
-bit value that causes the comparator output to become true
is selected as the A/D conversion value through a priority encoder. It is obvious that this type of
A/D converter will be very fast. However, they require a lot of hardware resources to implement
and therefore are not suitable for implementing high-resolution A/D converters. This type of A/D
converter is often used in applications that require high-speed but low resolution, such as a video
signal. Over the years, several variations to this approach have been proposed to produce high-
speed A/D converters. The most commonly used technique is to pipeline a flash A/D converter,
which will reduce the amount of hardware required while still achieving high conversion speed.
S
LOPE
AND
D
OUBLE
-S
LOPE
A/D C
ONVERTERS
This type of A/D converter is used in Microchip PIC14000 microcontrollers in which the
charging and discharging of a capacitor is used to perform A/D conversion. It requires relatively
simple hardware and is popular in low-speed applications, such as digital multimeters. In addi-
tion, high resolution (10- to 16-bit) can be achieved.
S
IGMA
-D
ELTA
A/D C
ONVERTERS
This type of A/D converter uses the
oversampling
technique to perform A/D conversion.
It has good noise immunity and can achieve high resolution. Sigma-delta A/D converters are
becoming more and more popular in implementing high-resolution A/D converters. The only
disadvantage is the slow conversion speed. However, this weakness is improving because of
advancements in CMOS technology.
S
UCCESSIVE
-A
PPROXIMATION
A/D C
ONVERTERS
The successive-approximation method approximates the analog signal to
n
-bit code in
n
steps. It may be used for low-frequency applications with large DC noise, such as an electro-
cardiograph. The block diagram of this method is shown in Figure 12.4.
It first initializes the successive-approximation register (SAR) to 0 and then performs a se-
ries of guessing, starting with the most-significant bit and proceeding toward the least-signifi-
cant bit. The algorithm of the successive-approximation method is illustrated in Figure 12.5.
For every bit of the SAR, the algorithm does the following operations:
•
Guesses the bit to be a 1
•
Converts the value of the SAR register to an analog voltage
•
Compares the D/A output with the analog input and clears the bit to 0 if the
D/A output is larger (which indicates that the guess is wrong)
Search WWH ::
Custom Search