Datasheet

2.0 Applications Information (Continued)
The minimum value of V
REF
(V
REF
=V
REF
+
−V
REF
−
) can be
quite small (see Typical Performance Characteristics) to al-
low direct conversion of transducer outputs providing less
than a 5V output span. Particular care must be taken with
regard to noise pickup, circuit layout and system error volt-
age sources when operating with a reduced span due to the
increased sensitivity of the converter (1 LSB equals V
REF
/
2
n
).
2.3 THE ANALOG INPUTS
Due to the sampling nature of the analog inputs, at the clock
edges short duration spikes of current will be seen on the
selected assigned negative input. Input bypass capacitors
should not be used if the source resistance is greater than
1kΩsince they will average the AC current and cause an
effective DC current to flow through the analog input source
resistance. An op amp RC active lowpass filter can provide
both impedance buffering and noise filtering should a high
impedance signal source be required. Bypass capacitors
may be used when the source impedance is very low without
any degradation in performance.
In a true differential input stage, a signal that is common to
both “+” and “−” inputs is cancelled. For the ADC10154 and
ADC10158, the positive input of a selected channel pair is
only sampled once before the start of a conversion during
the acquisition time (t
A
). The negative input needs to be
stable during the complete conversion sequence because it
is sampled before each decision in the SAR sequence.
Therefore, any AC common-mode signal present on the
analog inputs will not be completely cancelled and will cause
some conversion errors. For a sinusoid common-mode sig-
nal this error is:
V
error
(Max) = V
PEAK
(2πf
CM
)(t
C
)
where f
CM
is the frequency of the common-mode signal,
V
PEAK
is its peak voltage value, and t
C
is the A/D’s maximum
conversion time (t
C
= 22/f
CLK
for 10-bit plus sign resolution).
For example, for a 60 Hz common-mode signal to generate
a
1
⁄ 4 LSB error (1.24 mV) with a 4.5 µs conversion time, its
peak value would have to be approximately 731 mV.
2.4 OPTIONAL ADJUSTMENTS
2.4.1 Zero Error
The zero error of the A/D converter relates to the location of
the first riser of the transfer function (see
Figure 1 ) and can
be measured by grounding the minus input and applying a
small magnitude positive or negative voltage to the plus
input. Zero error is the difference between actual DC input
DS011225-21
a. Ratiometric Using the Internal Reference
DS011225-22
b. Absolute Using a 4.096V Span
FIGURE 5. Different Reference Configurations
ADC10154/ADC10158
www.national.com19