Do DHT11s "dry up" ?

Recently I bought a Si7021 temperature/relative humidity sensor and wrote a sketch that tells me what the sensor measures:
Current output of the Si7021: temp 24.2 Centigrade, rel humidity = 60%. This is the condition IN my little office.

Temperature is in line with what an analog mercury thermometer indicates.
Humidity is in agreement with what it feels, quite normal this type of weather, this time of the year (Netherlands, kind of summer, some showers today, some sunshine, semi-cloudy sky, currently evening, temperature dropping, clear sky).

I have two DHT11 temperature/relative humidity sensors bought one year ago. I constructed a weather station that measures with DHT11 #1 the temperature and relative humidity IN my little office (DHT11 mounted inside) and simultaneously with DHT11 #2 OUT of my office (DHT11 in a pod with punctured cap outside the window.and shielded from rain.
The IN DHT11 indicates 37% humidity (at 24 Celsius)and the OUT DHT11 presents 38% (at 18 Celsius) (humdity rising, temperature dropping) (which is OK if you live in dry Spain but not if you live in Holland)

Over the period of one year, the relative humidity indication gradually decreased. I am wondering whether my home and the environment of my home is drying up or whether DHT11's 'dry out' after some longer time.

Is there anyone in this forum who has experienced the same? Anyone who logged DHT11's over a long period.

Therefore, a forum topic: Do DHT11's suffer from degradatation of their relative humidity sensitivity.

Read the PDF documentation. Particularly, 7. Attentions of application. There are several ways to kill the sensor and some possible ways to reactivate it.


I have 2 of those sensors for over 1 year recording temp and humidity and plotting graphs in a Visual Basic program.

In my opinion, the DHT11 are not reliable at all. They give too low humidity readings and over this year operation, I saw no improvement or degradation. Many sites, on Internet, talk about the poor quality of those sensor.

If you want better results, I think you should go to more expensive device.

Thanks, gents for the comments. Getting accurate and reliable data out of an hygrometer has always been tricky. Remember the days with the hair hygrometer and the pine-cone model? Brr.

Which 'better' temp/rel humidity sensor to use? I quickly searched Aliexpress with search terms "humidity+sensor+arduino" and found these. I also tried to determine whether a library exist for that particular sensor type.

DHT11, DHT21 --------- library available? - yes
SHT 10, SHT15, SHT31 --------- library available? - yes
HS1101 --------- library available? - not that I know of
Si7021 --------- library available? - yes

The question is: which one is most accurate and has the longest life expectancy
I only have long-term experience with DHT11s (several with one and a half year continuous operation with apparent degradation - "drying up"). Indeed these sensors are very inaccurate
Recently I bought an Si7021 which works fine, however I have no prolonged user experience with this sensor. Does anybody have?

I am considering a small project wherein a DHT11, a Si7021 and a SHT are together soldered on a small board that will be connected to a logging device during the fall and winter. These seasons offer in the Low Countries usually a rapid sequence of cold, sometimes very dry spells and long periods with very wet conditions.

The wet and dry bulb thermometer is the old-fashioned method for measuring humidity and its
rather impractical to mimic electronically needing a constant supply of water!

There are bulky expensive spectrographic ways to measure water vapour directly, but the cheap
sensors all rely on some sort of partition coefficient of the water between the air and another medium,
such as the polymers used in the DHT11 style sensor. I suspect its hard to find a material that absorbs
water readily and gives it back readily and doesn't degrade long-term in the presence of oxygen, water and

Another humidity sensor is the AM2320. I use this one for a year now and within ±2% look like it's
doing the job.

However I have to use the following correction equation in order to have good lecture:

Humidity=Math.Round(((h_AM2320 - 11) ^ (1 / 2)) * 100) / 10)

Compare to a commercial meteo station in my room, I have similar results.