ML2252, ML2259
TYPICAL PERFORMANCE CURVES (Continued)
2
1.5
1
1
V
V
= 5V
= 0V
= 1.46MHz
= 25°C
CC
IN
V
= 5V
= 1.46MHz
CC
f
CLK
f
CLK
T
A
0.75
0.5
25°C
0.5
0
0.25
0
0
1
2
3
4
5
0
1
2
3
4
5
V
(V
REF DC
)
V
(V
REF DC
)
Figure 4. Unadjusted Offset Error vs V
Voltage
Figure 3. Linearity Error vs V
Voltage
REF
REF
1.0 FUNCTIONAL DESCRIPTION
1.2 A/D CONVERTER
1.1 MULTIPLEXER ADDRESSING
The A/D converter uses successive approximation to
perform the conversion. The converter is composed of the
successive approximation register, the DAC and the
comparator.
The ML2252 and ML2259 contain a single ended analog
multiplexer. A particular input channel is selected by
using the address decoder. The relationship between the
address inputs, ADDR0–ADDR2, and the analog input
selected is shown in Table 1. The address inputs are
latched into the decoder on the rising edge of the address
latch signal ALE.
The DAC generates the precise levels that determine the
linearity and accuracy of the conversion. The DAC is
composed of a capacitor upper array and a resistor lower
array. The capacitor upper array generates the 4 MSB
decision levels while the series resistor lower array
generates the 4 LSB decision levels. A switch decoder tree
is used to decode the proper level from both arrays.
ML2252
SELECTED
ANALOG CHANNEL
ADDRESS
INPUT
CH0
CH1
0
1
The capacitor/resistor array offers fast conversion, superior
linearity and accuracy since matching is only required
4
8
between 2 = 16 elements (as opposed to 2 = 256
elements in conventional designs). And since the levels are
based on the ratio of capacitors to capacitors and resistors to
resistors, the accuracy and long term stability of the
converter is improved. This also guarantees monotonicity
and no missing codes, as well as eliminating any linearity
temperature or power supply dependence.
ML2259
SELECTED
ANALOG CHANNEL
ADDRESS INPUT
ADDR1
ADDR2
ADDR0
CH0
CH1
CH2
CH3
CH4
CH5
CH6
CH7
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
The successive approximation register is a digital block
used to store the bit decisions from the conversion.
The comparator design is unique in that it is fully
differential and auto zeroed. The fully differential
architecture provides excellent noise immunity, excellent
power supply rejection, and wide common mode range. The
comparator is auto zeroed at the start of each conversion in
order to remove any DC offset and full scale gain error, thus
improving accuracy and linearity.
Table 1. Multiplexer Address Decoding
7
MICRO-LINEAR [ MICRO LINEAR CORPORATION ]
