Campbell CR10 PROM, Руководство оператора

Страница: 182 / 253

SECTION 13. CR10 MEASUREMENTS
13-3
In order to make a differential measurement, the
inputs must be within the CR10 common mode
range of
±
2.5 V. The common mode range is the
voltage range, relative to CR10 ground, within
which both inputs of a differential measurement
must lie in order for the differential measurement
to be made.
For example, if the high side of a differential input
is at 2 V and the low side is at 1 V relative to
CR10 ground, there is no problem; a
measurement made on the +2.5 V range would
indicate a signal of 1 V. However, if the high
input is at 2.8 V and the low input is at 2 V, the
measurement cannot be made because the high
input is outside of the common mode range. The
CR10 will indicate the overrange with the
maximum negative number (Section 3.5.)
Problems with exceeding common mode range
may be encountered when the CR10 is used to
read the output of external signal conditioning
circuitry if a good ground connection does not
exist between the external circuitry and the
CR10. When operating where AC power is
available, it is not always safe to assume that a
good ground connection exists through the AC
wiring. If a CR10 is used to measure the output
from a laboratory instrument (both plugged into
AC power and referencing ground to outlet
ground), it is best to run a ground wire between
the CR10 and the external circuitry. Even with
this ground connection, the ground potential of
the two instruments may not be at exactly the
same level, which is why a differential
measurement is desired (Section 7.2).
If a differential measurement is used on a sensor
that is not referenced to CR10 ground through a
separate connection (e.g., a net radiometer), a
jumper wire should be connected between the
low side of the differential input and analog
ground to hold the sensor in common mode
range.
A differential measurement has better noise
rejection than a single-ended measurement.
Integrating the signal in both directions also
reduces input offset voltage due to thermal
effects in the amplifier section of the CR10.
Input offset voltage on a single-ended
measurement is less than 5 microvolts; the
input offset voltage on a differential
measurement is less than 1 microvolt.
A single-ended measurement is quite
satisfactory in cases where noise is not a
problem and care is taken to avoid ground
potential problems. Channels are available for
twice as many single-ended measurements. A
single-ended measurement takes about half the
time of a differential measurement, which is
valuable in cases where rapid sampling is a
requirement.
NOTE: Sustained voltages in excess of
+16 VDC applied to the analog inputs will
damage the CR10 input circuitry.
13.3 THE EFFECT OF SENSOR LEAD
LENGTH ON THE SIGNAL SETTLING
TIME
Whenever an analog input is switched into the
CR10 measurement circuitry prior to making a
measurement, a finite amount of time is
required for the signal to stabilize at its correct
value. The rate at which the signal settles is
determined by the input settling time constant
which is a function of both the source
resistance, and input capacitance (explained
below). The CR10 allows a 450 µs settling time
before initiating the measurement. In most
applications this settling time is adequate, but
the additional wire capacitance associated with
long sensor leads can increase the settling time
constant to the point that measurement errors
may occur. There are three potential sources
of error which must settle before the
measurement is made:
1. The signal must rise to its correct value.
2. A small transient (
~
5 mV) caused by
switching the analog input into the
measurement circuitry must settle.
3. A larger transient, usually about 40 mV/V,
caused by the switched, precision
excitation voltage used in resistive bridge
measurements must settle.
The purpose of this section is to bring attention
to potential measurement errors caused when
the input settling time constant gets too large
and to discuss procedures whereby the effects
of lead length on the measurement can be
estimated. In addition, physical values are
given for three types of wire used in CSI
sensors, and error estimates for given lead
lengths are provided. Finally, techniques are