A paper, 'Measurement Errors Using a Geonor Weighing Gauge with a Campbell Scientific Datalogger', P2.5, was presented by H. Lamb and J. Swenson at AMS 2005 showing errors based on theory of 0.0012%reading/C in good agreement with the Geonor specification of 0.001%reading/C.
In the summers of 2005, 2007 and 2008, some 256 Geonor vibrating wire transducers were tested and aged in a Thermal Products Solutions (Lunaire/BlueM) ETCU-09 enviromental temperature/ humidity chamber in sets of 16 units, sub-sets of 4 units. Each sub-set of four supported a different fixed weight, 0.5 kg, 1.0 kg, 2.0 kg and 3.0 kg.
The results of this testing are presented, showing consistancy of performance and a negative temperature coefficient of about -0.004%reading/C which contrasts with the earlier value calculated from theory. The paper, P2.5, has been reviewed and an error found that led to a change in the sign of temperature coefficient. The theory actually indicates a value of -0.0012%reading/C.
Earlier tests on a gauge with one transducer and two supporting chains in an environmental chamber large enough to hold the entire unit show a more complicated picture, primarily due to changes in gauge leveling as temperature varied. Environment Canada has since upgraded the Geonors to triple transducers. Since the temperature coefficients are based on both the reading and temperature, it is possible, by using three correction equations, to produce a resultant gauge output which corrects not only for temperature variations but also for gauge level variations as a result of temperature.
The test results and the theory results are rationalized and temperature correction equations are presented.