Campbell Scientific CR1000 Measurement and Control System - Manual (Page 277)

Section 8. Operation
277
Sensors with a low signal-to-noise ratio, such as thermocouples, should normally
be measured differentially. However, if the measurement to be made does not
require high accuracy or precision, such as thermocouples measuring brush-fire
temperatures, a single-ended measurement may be appropriate. If sensors require
differential measurement, but adequate input channels are not available, an analog
multiplexer should be acquired to expand differential input capacity. Refer to the
appendix Analog Multiplexers
(p. 560)
for information concerning available
multiplexers.
Because a single-ended measurement is referenced to CR1000 ground, any
difference in ground potential between the sensor and the CR1000 will result in an
error in the measurement. For example, if the measuring junction of a copper-
constantan thermocouple being used to measure soil temperature is not insulated,
and the potential of earth ground is 1 mV greater at the sensor than at the point
where the CR1000 is grounded, the measured voltage will be 1 mV greater than
the true thermocouple output, or report a temperature that is approximately 25°C
too high. A common problem with ground-potential difference occurs in
applications wherein external signal conditioning circuitry is powered by the same
source as the CR1000, such as an ac mains power receptacle. Despite being tied
to the same ground, differences in current drain and lead resistance may result in a
different ground potential between the two instruments. Hence, a differential
measurement should be made on the analog output from an external signal
conditioner. Differential measurements MUST be used when the low input is
known to be different from ground.
8.1.2.3 Measurement Sequence
The CR1000 measures analog voltage by integrating the input signal for a fixed
duration and then holding the integrated value during the successive
approximation analog-to-digital (A/D) conversion. The CR1000 can make and
store measurements from up to 8 differential or 16 single-ended channels at the
minimum scan interval of 10 ms (frequency of 100 Hz) using fast-measurement-
programming techniques as discussed in Fast Measurement Rates
(p. 231).
The
maximum conversion rate is 2000 per second (2 kHz) for measurements made on
a single channel.
The timing of CR1000 measurements is precisely controlled. The measurement
schedule is determined at compile time and loaded into memory. This schedule
sets interrupts that drive the measurement task.
Using two different voltage measurement instructions with the same voltage range
takes the same measurement time as using one instruction with two repetitions.
Note This is not the case with legacy CR10(X), 21X, CR23X, and CR7(X)
dataloggers. Using multiple measurement "reps" in these dataloggers reduced
overall measurement time.
Several parameters in CRBasic voltage measurement instructions VoltDiff() and
VoltSE() vary the sequence and timing of measurements. Table CRBasic
Parameters Varying Measurement Sequence and Timing
(p. 278)
lists these
parameters.