Calibration of hygrometers at non-static conditions

• Published in: Measurement science & technology Vol. 31; no. 3; pp. 34003 - 34013
• Main Authors: Högström, R, Salminen, J, Heinonen, M
• Format: Journal Article
• Language: English
• Published: IOP Publishing 01.03.2020
• Subjects: calibration; hygrometer; relative humidity
• ISSN: 0957-0233; 1361-6501
• DOI: 10.1088/1361-6501/ab56a6

The current practice of calibrating humidity sensors at discrete measurement points under static conditions is time-consuming, i.e. expensive, due to the long stabilization times. In order to reduce the calibration time, we introduce a new calibration method based on humidity ramp measurements, i.e. measurements are performed at first while humidity is increased with constant speed and then decreasing with the same speed. A calibration system based on the mixing flow humidity generation principle was developed for generating linear humidity ramps inside a measurement chamber. Two different calibration approaches were investigated. In the first one, high accuracy was targeted with moderate ramp speeds (2 h to 12 h) and a large volume measurement chamber using a chilled mirror hygrometer as a reference. In the second approach, a shorter calibration time was achieved with fast ramp speeds (0.5 h to 2 h) and a small volume chamber using a fast capacitive humidity sensor as a reference. The developed calibration procedure was validated by comparing results of non-static and static calibrations to each other for five different capacitive humidity sensors from two manufacturers. The non-static calibration was found to provide equivalent results compared to the static calibration with a potential of reducing the overall calibration time by up to 50%. Preconditions of the non-static calibration related to the ramp speed and sensors response times are discussed, and an estimation of the calibration uncertainty is given. The main advantage of the developed non-static calibration method is that calibrations can be performed faster and more data on the sensor behaviour is obtained than with the conventional point-wise calibration without any significant increase in uncertainty.