LTC2410
26
APPLICATIO S I FOR ATIO
WU
U
For relatively small values of input capacitance (CIN <
0.01
F), the voltage on the sampling capacitor settles
almost completely and relatively large values for the
source impedance result in only small errors. Such values
for CIN will deteriorate the converter offset and gain
performance without significant benefits of signal filtering
and the user is advised to avoid them. Nevertheless, when
small values of CIN are unavoidably present as parasitics
of input multiplexers, wires, connectors or sensors, the
LTC2410 can maintain its exceptional accuracy while
operating with relative large values of source resistance as
shown in Figures 17 and 18. These measured results may
be slightly different from the first order approximation
suggested earlier because they include the effect of the
actual second order input network together with the non-
linear settling process of the input amplifiers. For small CIN
values, the settling on IN+ and IN– occurs almost indepen-
dently and there is little benefit in trying to match the
source impedance for the two pins.
Larger values of input capacitors (CIN > 0.01F) may be
required in certain configurations for antialiasing or gen-
eral input signal filtering. Such capacitors will average the
input sampling charge and the external source resistance
will see a quasi constant input differential impedance.
When FO = LOW (internal oscillator and 60Hz notch), the
typical differential input resistance is 1.8M
which will
generate a gain error of approximately 0.28ppm for each
ohm of source resistance driving IN+ or IN–. When FO =
HIGH (internal oscillator and 50Hz notch), the typical
differential input resistance is 2.16M
whichwillgenerate
a gain error of approximately 0.23ppm for each ohm of
source resistance driving IN+ or IN–. When FO is driven by
an external oscillator with a frequency fEOSC (external
conversion clock operation), the typical differential input
resistance is 0.28 1012/fEOSC and each ohm of
source resistance driving IN+ or IN– will result in
1.78 10–6 fEOSCppm gain error. The effect of the source
resistance on the two input pins is additive with respect to
this gain error. The typical +FS and –FS errors as a function
of the sum of the source resistance seen by IN+ and IN– for
large values of CIN are shown in Figures 19 and 20.
In addition to this gain error, an offset error term may also
appear. The offset error is proportional with the mismatch
between the source impedance driving the two input pins
IN+ and IN– and with the difference between the input and
reference common mode voltages. While the input drive
circuit nonzero source impedance combined with the
converter average input current will not degrade the INL
performance, indirect distortion may result from the modu-
lation of the offset error by the common mode component
of the input signal. Thus, when using large CIN capacitor
values, it is advisable to carefully match the source imped-
ance seen by the IN+ and IN– pins. When FO = LOW
(internal oscillator and 60Hz notch), every 1
mismatch
in source impedance transforms a full-scale common
mode input signal into a differential mode input signal of
0.28ppm. When FO = HIGH (internal oscillator and 50Hz
notch), every 1
mismatch in source impedance trans-
forms a full-scale common mode input signal into a
differential mode input signal of 0.23ppm. When FO is
driven by an external oscillator with a frequency fEOSC,
every 1
mismatch in source impedance transforms a
full-scale common mode input signal into a differential
mode input signal of 1.78 10–6 fEOSCppm. Figure 21
shows the typical offset error due to input common mode
voltage for various values of source resistance imbalance
between the IN+ and IN– pins when large CIN values are
used.
If possible, it is desirable to operate with the input signal
common mode voltage very close to the reference signal
common mode voltage as is the case in the ratiometric
measurement of a symmetric bridge. This configuration
eliminates the offset error caused by mismatched source
impedances.
The magnitude of the dynamic input current depends upon
the size of the very stable internal sampling capacitors and
upon the accuracy of the converter sampling clock. The
accuracy of the internal clock over the entire temperature
and power supply range is typical better than 0.5%. Such
a specification can also be easily achieved by an external
clock. When relatively stable resistors (50ppm/
°C) are
used for the external source impedance seen by IN+ and
IN–, the expected drift of the dynamic current, offset and
gain errors will be insignificant (about 1% of their respec-
tive values over the entire temperature and voltage range).
Even for the most stringent applications, a one-time
calibration operation may be sufficient.