YSI 63 - Manual (Page 27)

23
Temperature Error = ± 1% maximum
The solution temperature error is the product of the temperature coefficient and the temperature
offset from 25°C, expressed as a percentage of the reading that would have been obtained at
25°C. The error is not necessarily a linear function of temperature. The statement of error is
derived from a 25°C temperature offset and a 3%/°C temperature coefficient.
Total Error
Considering only the above three factors, system accuracy under worst case conditions will be
±2%, although the actual error will be considerably less if recommended and properly calibrated
cells and instrument ranges are used. Additional errors, which can essentially be eliminated with
proper handling, are described below.
Cell Contamination
This error is usually due to contamination of the solution being measured, which occurs when
solution is carried-over from the last solution measured. Thus, the instrument might be correctly
reporting the conductivity seen, but the reading does not accurately represent the value of the
bulk solution. Errors will be most serious when low conductivity solutions are contaminated by
carry-over from high conductivity solutions, and can then be of an order of magnitude or more.
Follow the cleaning instructions carefully before attempting low conductivity measurements
with a cell of unknown history or one that has been previously used in higher value solutions.
An entirely different form of contamination sometimes occurs due to a buildup of foreign
material directly on cell electrodes. While rare, such deposits have, on occasion, markedly
reduced the effectiveness of the electrodes. The result is an erroneously low conductance
reading.
Electrical-Noise Errors
Electrical noise can be a problem in any measurement range, but will contribute the most error
and be the most difficult to eliminate when operating in the lowest ranges. The noise may be
either line-conducted or radiated or both, and may require, grounding, shielding, or both.
Galvanic and Miscellaneous Effects
In addition to the error sources described above, there is another class of contributors that can be
ignored for all but the most meticulous of laboratory measurements. These errors are always
small and are generally completely masked by the error budget for cell-constant calibration,
instrument accuracy, etc. Examples range from parasitic reactances associated with the solution
container and its proximity to external objects to the minor galvanic effects resulting from oxide
formation or deposition on electrodes. Only trial and error in the actual measurement
environment can be suggested as an approach to reduce such errors. If the reading does not
change as the setup is adjusted, errors due to such factors can be considered too small to see.