Specifications

Section 3. CR3000 Measurement Details
used ± (0.X% of Full-Scale Range) to specify measurement accuracy for a given
range, which results in a fixed number for each input range.
While the ± (0.X% of Full-Scale Range) method is simple to test and utilize in
accuracy estimates, it overestimates errors for inputs much less than full-scale.
This is because typical measurement instrumentation has measurement errors
composed of two major components; a gain error and an offset error as illustrated
in Figure 3.9-1. Gain error is minimized by factory calibration, but tends to
increase with temperature deviations and/or component aging. An accuracy
specification of ± (0.X% of reading + Offset) better represents the instrument
performance illustrated in Figure 3.9-1 than does ± (0.X% of Full-Scale Range).
Campbell Scientific, Inc. utilizes ± (0.X% of reading + Offset) accuracy
specifications for the CR9000, CR5000, CR3000 and CR1000 dataloggers.
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
-3000 -2000 -1000 0 1000 2000 3000
Input Voltage (mV)
Voltage Measurement Error (mV)
FIGURE 3.9-1. Instrument voltage measurement error performance illustrating
dependence upon input voltage.
The following discussion relates to the CR10X and CR1000 dataloggers, but the
principles are applicable to the CR3000 as well.
A graphical representation of the CR10X ± (0.X% of Full-Scale Range) and
CR1000 ± (0.X% of reading + Offset) voltage measurement accuracy
specifications is illustrated in Figure 3.9-2.
3-23