The BME680 needs to run for some time to stabilize to its environment. Generally, the longer a BME680 was turned off, the longer it may take to stabilize, but we estimate these 5 minutes operation in LP mode to be typically sufficient. Without any information, BSEC must assume that it has just been started for the first time and possibly for a sensor that hasn't been running for a long time, therefore BSEC starts all algorithms and processes from scratch, including this 5min period.
@LeGuépard wrote:
Just another question, is it always necessary to leave the BME680 working for more than 5 mins, in LP mode, before having the real readings on the IAQ ?, or is it just needed for the first time when someone uses the BSEC ?
Yes due to the above, but..
Optionally, BSEC supports a state file to be saved to some external memory which can be later restored. This feature is handful for example if you must replace the battery of a device without loosing BSEC's self-calibration history. In case you were to reload a valid state file after having replaced the battery, you would see that the IAQ values would already be changing in the first few minutes. But since the sensor is still affected by the behavior described above, there is a low confidence on the data quality whithin this period that is still indicated by an IAQ accuracy of zero as a result.
@LeGuépard wrote:
And during those 5 mins, how accurate are the other variables (temperature, humidity, pressure. etc.) ?
When a board or device starts running, it may take a bit of time for the pcb and its components to stabilize to new ambient conditions as well, thus slightly affecting the temperature and relative humidity measurements, but the effect is typically not significant for the pressure output.
... View more