In a room with 26.3 degrees Celcius TEMP1 outside its casing at the same location as two other sensors measures 5 degrees more on the onboard chip. The humidity shows 12% less. Actually is seems the onboard temp is more or less the CPU temp (making the on board sensor useless) und the humidity simply is highly incurrat. Not very convincing.
Calibration in HA is a problem, as it seems it only allows for a fixed absolut offset. From what i have seen so far, the offset changes based on the env temp. So calibration only shows to work on the temp the claribration is set.
Bottom line it seems highly integrated boards are the better choice than ESP based boards with sensors, as the ESP obv. is running too hot to not interfere with the sensor.
In a room with 26.3 degrees Celcius TEMP1 outside its casing at the same location as two other sensors measures 5 degrees more on the onboard chip. The humidity shows 12% less. Actually is seems the onboard temp is more or less the CPU temp (making the on board sensor useless) und the humidity simply is highly incurrat. Not very convincing.
Calibration in HA is a problem, as it seems it only allows for a fixed absolut offset. From what i have seen so far, the offset changes based on the env temp. So calibration only shows to work on the temp the claribration is set.
Bottom line it seems highly integrated boards are the better choice than ESP based boards with sensors, as the ESP obv. is running too hot to not interfere with the sensor.