hi! I bought myself an esp01 equipped with dht22 temp/hum sensor. flashed it with espeasy and are using the hum% for bathroomfan regulation. The sensor seems to be very inaccurate. Hum rise from 30 to 50% for a minute or so than back to 30.
This kills the proper working of the script. see
https://www.domoticz.com/wiki/Humidity_control
Question: what's a reliable sensor? or is there a way to correct sensorvalue or just try another dht22.
Any suggestion is welcome. Thanks!
Bart
accuracy dht22 sensor
Moderators: grovkillen, Stuntteam, TD-er
Re: accuracy dht22 sensor
According to this page the DHT22 should be quite OK, but maybe it suffers condensation where it is currently located, and is disturbed by that.
Reading that list above, it seems the ATH20 is a better choice, as it has a range of 0..100%RH, where the DHT22 only allows 5..95%RH
In another story I read (no link) about an outdoor placed (in a proper housing) BME680, that also allows 0..100%RH, that had been working reliably for a year.
Unfortunately you are using a 1MB ESP (most likely, the default for an ESP01S), so you won't be able to install the Climate or Collection A builds, as those have the AHT10/AHT20 plugin included. But you can replace the flash with a 4MB chip, that makes installing these ESPEasy builds possible
Or if the ESP01 is in a socket, you can replace it with the T-01C3esp32 that has 4MB flash and the supported ESP32-C3 chip (plug & play replacement, just different ESPEasy build)
Reading that list above, it seems the ATH20 is a better choice, as it has a range of 0..100%RH, where the DHT22 only allows 5..95%RH
In another story I read (no link) about an outdoor placed (in a proper housing) BME680, that also allows 0..100%RH, that had been working reliably for a year.
Unfortunately you are using a 1MB ESP (most likely, the default for an ESP01S), so you won't be able to install the Climate or Collection A builds, as those have the AHT10/AHT20 plugin included. But you can replace the flash with a 4MB chip, that makes installing these ESPEasy builds possible
Or if the ESP01 is in a socket, you can replace it with the T-01C3esp32 that has 4MB flash and the supported ESP32-C3 chip (plug & play replacement, just different ESPEasy build)
/Ton (PayPal.me)
Re: accuracy dht22 sensor
I know some DHTxx sensors (not sure which) could get "locked" into some bogus state when they get saturated.
You then need to keep them in a warm and extremely dry place for a while to get working again. (e.g. in uncooked rise in the sun or near some food dehydration device)
BME280 is probably also fine for this.
As for any temperture/humidity sensor, just make sure it isn't placed in the same enclosure as the ESP or else it might heat up and give unreliable results.
For the BME280 I have included some temperature offset selector which also corrects the humidity accordingly.
You then need to keep them in a warm and extremely dry place for a while to get working again. (e.g. in uncooked rise in the sun or near some food dehydration device)
BME280 is probably also fine for this.
As for any temperture/humidity sensor, just make sure it isn't placed in the same enclosure as the ESP or else it might heat up and give unreliable results.
For the BME280 I have included some temperature offset selector which also corrects the humidity accordingly.
Re: accuracy dht22 sensor
Ath, TD-er,
Thanks for replies. I am using the assembly esp01 with dht22.
Ventilation is arranged with 125mm pvc tube through wall direct to outside. No return valve included. The fan is a tubefan mounted inside the pvc tube. The esp assembly jusf behind the grid in bathroom.
My first try will be to have a nodemcu or wemos with 2 or 3 sensors nearby the shower to see what will be the output of these DHT's . I also have some DHT's 11. Correct measuring is less important. But temporarily 'strange values' for over one minute is killing the automated controls.
Might there be a way to have the average of some sensors to overcome faulty measures. E.g. 3 sensors of which 1 sensor deviates too much from both others? This is just a thought.
Bart
Thanks for replies. I am using the assembly esp01 with dht22.
Ventilation is arranged with 125mm pvc tube through wall direct to outside. No return valve included. The fan is a tubefan mounted inside the pvc tube. The esp assembly jusf behind the grid in bathroom.
My first try will be to have a nodemcu or wemos with 2 or 3 sensors nearby the shower to see what will be the output of these DHT's . I also have some DHT's 11. Correct measuring is less important. But temporarily 'strange values' for over one minute is killing the automated controls.
Might there be a way to have the average of some sensors to overcome faulty measures. E.g. 3 sensors of which 1 sensor deviates too much from both others? This is just a thought.
Bart
Re: accuracy dht22 sensor
That plugin has the plugin statistics enabled, but that entire feature is disabled for the Collection, Minimal and 1M (like the ESP01S) builds (and other 'bin-size limited' ESP8266 builds)... 
When buying new ESP's go for an ESP32 (dual-core, stable workhorse), ESP32-S3 (dual-core, many GPIO's) or ESP32-C3 (single-core, ESP8266 successor, small boards available, search for 'ESP32 C3 mini'), the ESP32 and ESP32-S3 are also widely available with 16MB Flash, so you can install a MAX build, including all plugins and features.
When buying new ESP's go for an ESP32 (dual-core, stable workhorse), ESP32-S3 (dual-core, many GPIO's) or ESP32-C3 (single-core, ESP8266 successor, small boards available, search for 'ESP32 C3 mini'), the ESP32 and ESP32-S3 are also widely available with 16MB Flash, so you can install a MAX build, including all plugins and features.
/Ton (PayPal.me)
Who is online
Users browsing this forum: No registered users and 1 guest