Accuracy query for DHT22 users?

Hi All - wondering what kind of accuracy and repeatability users are seeing with DHT22 temp/humidity sensors?

My temperature reading seems to be pretty good but I am not sure about the RH I am getting

What are others experiencing - especially at the LOW end - I feel my is reading high by a few percent at the bottom of the scale (ie when gov sites indicate 5-7%, mine seems to indicate 10-12%)

Through the middle and high ranges it seems pretty good - its just a the low end

I am not sure if its my sensor (waiting on more stock to test this theory) or some environmental factor at my test site.


I don't use the DHT11 or DHT22 anymore. The DHT11 could be off by 50% and the DTH22 by 20 or 30%.

Storing those sensors in the wrong type of plastic bag could harm the sensor a lot, or perhaps the ones that I bought from Ebay are not calibrated/initialized. I have not done any calibration/initialization myself yet, it is not easy.

May I ask what you are using for RH instead?

I have 2 DHT22's and they both give almost the same temperature and RH, always. I can't say how accurate they are because I don't have any GOOD thermometer/hygrometer to compare.

I think that for the cost, they are pretty good :)

I have only used the DHT11. Temperature only has a resolution of 1 degree. The one I used consistently reports 1 degree above actual. An adjustment of -1 would make it fairly accurate but I've never used it for it's temperature reading. On the humidity scale it seems to be within about 3% (comparing with other more expensive kit) Consistently about 3% under. So an adjustment is fairly simple for this too. There are not many applications where humidity has to be controlled any finer than this range, so I think it's an ideal tool for the job. Especially at the low cost that these sell for.

I feel my is reading high

What standard are you basing this feeling off of? More stock isn't going to increase accuracy or help you determine how close you are to the actual value, though it might give you a little more of a normal distribution. Honestly, you get what you pay for. I wouldn't expect excellent accuracy from a $4-$5 part. However, in my experience, the DHT sensors work very well and reliably. I feel 2%-5% accuracy is acceptable for the price paid.

You might find this link interesting:

My comparison source consists of a local Government operated weather station and a couple of nearby Davis weather stations. My "feeling" is only after a couple of days field operation so i will see how its tracking after a week.

Anyone else have any feedback about how they have found the DHT sensors at very low RH values?

First, a shameless plug for my own web pages: From there you can find links to all the tests I am running on the DHT22 as well as DHT11 and far more expensive SHT71. The most direct answer to your question would be this plot

To more explicitly answer your question, I would agree with the level of errors you are getting as being typical. Most polymer capacitance hygrometer data sheet you look at will show larger errors below 20% and above 90%. Those extremes are always difficult to measure. Most of the sensors I have tested read about 4% higher than expected at very low humidity. The systematic consistency of all my sensors reading high does leave me a little concerned that my calibration technique could be flawed, but there could be lots of good reasons reasons for it being true. In any case, my personal experience comparing six DHT22s suggest that 5% would be a fair estimate of accuracy. Some are better than that.

I would say the numbers you quote are very close to my experience for this sensor's accuracy, but you are also right that there could be local environmental factors making your site differ from the comparison.

@rjs3273 Great work!!!

I added a link to your to my library (although I do not know if you used it)

See - -

If it needs additional info please let me know (PM + refer to this post)