Kipp&Zonen CFR Calibration Facility - Manual (Page 18)

CFR manual
17
3.3.3 Voltmeter offset
Theoretically, the voltmeter can cause a zero offset or drift. Short-circuiting the voltmeter input with a
resistance equal to the pyranometer’s impedance can check the zero offset. The voltmeter might change
sensitivity over the years. Generally it should be regularly recalibrated, as for all test and measurement
instruments. Because of the fact that the calibration is a relative measurement, the voltmeter sensitivity is
cancelled out of the equations. Only the voltmeter linearity and short term drift must be good.
3.3.4 Tilting of the pyranometers
Tilting of the pyranometers, and also the related possibility of light rays that fall in at an angle other than
perpendicular, is not a significant error source. This can easily be shown by comparing the cosine values of
0° and 3° angles. These differ by only 0.1%.
3.3.5 Differences in sensor height and geometry
As stated earlier in this chapter, the main purpose of the calibration procedure is to perform a one-to-one
comparison of the reference pyranometer and the test pyranometer, which should ideally be of the same
model. In that case the height and geometry of both the reference and the test pyranometers are exactly
the same. In general, errors in height are the most serious threats to measurement accuracy and should be
checked with each calibration.
Note that for side-by-side calibrations outdoors the sensor’s heights are less critical because the light
source )the sun) is distant and the intensity gradient is small.
3.3.6 Calibration repeatability and overall uncertainty
From experience, the agreement of the calculated values of sensitivity (S) over a number of consecutive
measurements of the same test pyranometer shows repeatability within 0.5% percent.
The overall calibration uncertainty of the test pyranometer is estimated to be < 1% compared with the
calibrated value of a reference Pyranometer of similar type.