I have a BME280 in an animal enclosure, monitoring temperature and relative humidity.
About 8 months ago, I noted that the RH was reading way lower than I had expected, but had assumed that it was correct. I changed out the sensor, and the readings went straight back to what I thought it should be. Last week, I noticed the issue again, but this time I had it trended. Unfortunately, it is still outside the history range (to the the left of the graph), but what happened was that the humidity plummeted on a step change from about 70% down to 40% and consistently read 30% low from then on. This can be seen as it hovers around 40% in the graph below. (the straigh horizontal lines were a loss of data connection) There was no obvious reason for the step change.
I shut down the system down last night and restarted it to get the first vertical line step change seen on the right of the graph, from 40% down to near 10%, which climbed over night to about 35% (about half of what it really is). With this poor performance, I swapped this sensor out with the original sensor that malfunctioned about 6 months ago, and the readings popped right back up to where it should be.
So what is happening here? The sensor is mechanically protected, all the connections are heat shrunk and terminated properly. The only abnormal thing is that the sensor is in a relatively high humidity environment constantly. Am I operating the sensor out it's specification or what?
Any input would be appreciated.
Talked to our internal experts, they need get more information.
Only based on the graph you provided, we knew what happened, but it is hard to find the reason why this situation happened.
could you provide the information of testing environment, what happened during this 8 months? Meanwhile, Could you also offer the reference device value of this 8 months ?
The sensor is mounted on a break out board, which is this one commonly seen in the arduino/ESP/IOT community. The sensor is in a plastic enclosure (basically a pipe about 15mm ID) and protected on 3 sides and connected to an ESP32 using an I2C interface.
Humidity is constant around 60% RH and the temperature cycles daily between 23 and 27 to simulate a coastal tropical environment. I believe the sensor is shielded from water, but one can never be certain if a drop of condensation has ever gotten to it.
But the two times this has happened, it was a step change, and it never went back until changing the sensor, however, the first sensor 'righted' itself since i reinstalled it when the 2nd sensor went bad the same way.
That's all I can really think of that describes it's environment and usage. Powering off and on does not seem to rectify the issue.
Not only it's environment of sensor place, the real time data of temperture and humidity also need be offered, include reference device values/ sensor data.
Otherwise, if no sensor data and reference values, we don't know what happened and have no way to do further analysis.
Sorry to say these, thanks for your understanding.