Rupiert
Posts: 13
Joined: Thu Apr 04, 2019 8:00 pm

DHT22 Accuraty.

Sat Apr 27, 2019 4:23 am

Hey there!

I have got a DHT22 sensor in a humidor. :D I put it next to a moisture meter. I don't have any information about accuracy for the moisture meter. Both display the same temperature. If the humidity raises the Sensor is +40% upper than the moisture meter. So the moisture meter shows me 64% and the sensor partly 90%. I know out of the Datasheet of the DHT22, it can vary +-5% RH. Can the Sensor be right at this high level??? What can I do? :idea:

Something awful like: :cry:

Code: Select all

if DHT22.humanity > 50 {
    if DHT22.humanity > 60 {
        if DHT22.humanity > 70 {
            if DHT22.humanity > 80 {
                if DHT22.humanity > 90 {
                    humanity = humanity - 10
                }
                humanity = humanity - 10
            }
            humanity = humanity - 10
        }
        humanity = humanity - 10
    }
    humanity = humanity - 10
}

User avatar
Joel_Mckay
Posts: 288
Joined: Mon Nov 12, 2012 10:22 pm
Contact: Website

Re: DHT22 Accuraty.

Sat Apr 27, 2019 6:11 am

We use the BME280 (several years of uptime), and I dumped the DHT/SHT units we sampled in the e-waste recycling bucket after testing the repeatability over several months (cleaved the leads off to ensure they stayed gone).

I suggest you do the same... as although there will be a few people who insist all DHT models work reliably without pinning the cpu in a busy loop...
That was not what the 300 unit test-samples we had showed, as the sensor itself tended to randomly latch-up after a few weeks. I observed this failure mode in several variants where they required a complete power on-off cycle to reset the DHT internal logic. The DHT variants would then continue to behave for a few more weeks under the identical operating conditions... then bork again... This reset behavior fools people into thinking they fixed it after doing some tweak... but that is rarely the case.

Also, in terms of accuracy others confirmed the bme280 ranks high out of 11 different experimentally verified models:
http://www.kandrsmith.org/RJS/Misc/Hygr ... _many.html

Please refer people to the above URL in the future, as it will save other students some time getting stuff to work right.

Cheers,
j

MarkDH102
Posts: 326
Joined: Fri Feb 13, 2015 3:18 pm

Re: DHT22 Accuraty.

Sat Apr 27, 2019 6:15 am

I too use the BME280.
I used to use the DHT22 but running the 2 in parallel showed the BME280 seemed to be more accurate.
My DHT22 were always reading about 20% over the correct level - but were consistent...

Return to “Troubleshooting”