Accuracy that lasts: an analysis of humidity sensor drift

How often should I calibrate my humidity sensors and how much drift can I expect from them? These are two common questions we are asked, but they are very difficult to answer without more information as the conditions where the sensors are used have a big impact on their stability.

Here is an attempt to give you some insight.

Our HMP110 humidity probes are used as both standalone humidity and temperature transmitters and as the measurement engine in our HMT120 and HMT130 humidity transmitters. Many of our customers also trust us with the calibration of their HMP110 probes. 

Image

We have analyzed the as-found results from service calibrations performed in our Helsinki, Boston, Tokyo, and Beijing service centers. Although we have no information on how the probes have been used we assume that most of them are used to monitor conditions in cleanrooms or datacenters, which are typically not very demanding applications for these products. We analyzed about 5,000 calibrations for 2,000 individual HMP110 probes, with some probes being calibrated multiple times. We assumed that no other adjustments had been performed during the lifetime of the probes. The as-found RH errors at 0%RH and 75%RH, as well as the ambient temperature measurement, were plotted against the time since the previous calibration:

Image

Image

After two years the main statistics for the as-found errors are as follows:

(Extreme outliers have been removed, leaving 170samples. 

Without going too deep into the statistical analysis, our main conclusions are that:

• most of the probes have remained within their as-new specification (±1.5%RH/0.1 °C) even after two years of use 

• the drift estimates for uncertainty calculations of ±0.5%RH/year and ±0.02 °C/year seem fair  

Learn more about HMP110, the trouble-free and cost-effective humidity transmitter with high accuracy and good stability.