DHT11 and DHT22 testing

​The DHT11 sensor that comes with the Higrow boards are extremely inaccurate, but how much better are the more expensive DHT22 (aka AM2302)?

​I’m now updating my Higrow boards to use DHT22/AM2302 over the DHT11 sensors that come with the boards, but how precise are these sensors? I tested three of each sensor. The sensors I tested are from different purchases over the last two years. They all look official, but I have no idea how the original is supposed to be so I can only guess that they are all valid. LastMinuteEngineers (great name!) have a good overview of specs and how to use these.

The sensors are quite similar, but the datasheets promise a higher precision for temperature (+/-0.5 Celcius vs +/-2.0 Celcius) and a wider range & precision for humidity. To test the sensors, I made a simple setup (as in the picture) with 3 sensors next to each other. I first tested the three DHT22/AM2302 and then the DHT11. The test setup stood still for about 15 minutes so that all sensors could absorb moisture in the air. I then averaged 50 samplings for each sensor. These are the results:

	Temperature	Humidity
AM2302 (A)	27.50 *C	49.90%
AM2302 (B)	27.60 *C	59.90%
AM2302 (C)	27.90 *C	34.40%
DHT11 (A)	26.00 *C	36.00%
DHT11 (B)	29.00 *C	43.00%
DHT11 (C)	27.00 *C	20.00%

As can be seen, the DHT11’s are highly imprecise, but the temperature measurements for the AM2302 show that these are quite solid and repeatable. For measuring moisture, neither of them are very good. The AM2302’s are closer matched, but far from the +/- 5% promised in the datasheets.