DHT-22 sensor outages in use with Homeduino
Hey folks,
I would like to ask if someone else got same behavior with DHT-22 temp/humidity sensors connected to an Arduino Nano: under unknown circumstances it appears to have a malfunction in reading the right values that spoils the complete graph view 
Regards and a happy new Year for this awesome project,
Paule
DHT ,DS18B20 and other I2C Devices submitting often wrong values over the bus.
You have to check in your sketch if the value is correct/valid
you can use the pimatic filter plugin for that https://pimatic.org/plugins/pimatic-filter/
pimatic v0.9 has been released!
Support Pimatic and get some free stickers
Like us on Facebook
make it so !
It may also help to use a capacitor to filter Vc noise and to terminate the 1wire bus as recommended.
Besides this, there is a ‘dhtlib-update’ branch of the homeduino sketch with an updated version of the DHT library which you can trial. See https://forum.pimatic.org/topic/2392/dht22-extreme-values for details
"It always takes longer than you expect, even when you take into account Hofstadter's Law.", Hofstadter's Law
@mwittig said in DHT-22 sensor outages in use with Homeduino:
It may also help to use a capacitor to filter Vc noise and to terminate the 1wire bus as recommended.
Thank you for your reply.
So what is your recommendation about the capacitor (µF) regardinig noise and where should i place it (powersupply or dataline to DHT-22)? there is already an 10k resistor between the 5Vs and the data line…
btw: there is also another DHT-22 running without any outages! possibly another Hardware rev.?! one is connected to PIN12 of Arduino and the other one to PIN11…
@Cybolord said in DHT-22 sensor outages in use with Homeduino:
So what is your recommendation about the capacitor (µF) regardinig noise and where should i place it (powersupply or dataline to DHT-22)? there is already an 10k resistor between the 3v3 and the data line…
My recommedation is to use a 100nF capacitor between Vcc and GND as sketched in the data sheet http://www.micropik.com/PDF/dht11.pdf. There are other sources suggesting other resistance values, but I haven’t tried that. The geneneral purpose of the capacitor is to smooth the supply on demand and to decouple high frequency signals, noise, from entering or leaving the circuit.
As far as I know the pull-up should be 4.7k assuming a cable length of less than 1m. For longer cable lengths it must be even lower. The 10k may be ok, though.
btw: there is also another DHT-22 running without any outages! possibly another Hardware rev.?! one is connected to PIN12 of Arduino and the other one to PIN11…
The DHT-22 is a different circuit than DHT-11. It should also be said that there are really cheap makes of DHT sensors being offered which may have quality issues. If you paid less than 1 EUR for your DHT-11 sensor you probably should get another one for testing …
My recommendation is not to use DHT-xx sensor unless you have good reasons to do so (e.g. you want use 1wire for infouse cabling of sensors). Personally, I have made good experiences with SI7020 I2C sensors.
"It always takes longer than you expect, even when you take into account Hofstadter's Law.", Hofstadter's Law
Perfect, thank you for your quick reply Marcus!
@mwittig said in DHT-22 sensor outages in use with Homeduino:
My recommedation is to use a 100nF capacitor between Vcc and GND as sketched in the data sheet http://www.micropik.com/PDF/dht11.pdf. There are other sources suggesting other resistance values, but I haven’t tried that. The geneneral purpose of the capacitor is to smooth the supply on demand and to decouple high frequency signals, noise, from entering or leaving the circuit.
I guess this has to be a ceramik type of capacitor?! will try that…
As far as I know the pull-up should be 4.7k assuming a cable length of less than 1m. For longer cable lengths it must be even lower. The 10k may be ok, though.
one DHT-22 is round about 5meters away; the other DHT-22 is round about 9meters away - so I’ll unsolder the 10k res and will replace it with 4.7k res - thx for the hint
The DHT-22 is a different circuit than DHT-11. It should also be said that there are really cheap makes of DHT sensors being offered which may have quality issues. If you paid less than 1 EUR for your DHT-11 sensor you probably should get another one for testing …
my fault, have to clarify: this is NO DHT-11 sensor used at all - only DHT-22 types - one for about EUR 6,50 sent from Belgium and the other one for about EUR 2,50 sent from China so you may be right…
@Cybolord said in DHT-22 sensor outages in use with Homeduino:
I guess this has to be a ceramik type of capacitor?! will try that…
Yes.
one DHT-22 is round about 5meters away; the other DHT-22 is round about 9meters away - so I’ll unsolder the 10k res and will replace it with 4.7k res - thx for the hint
It may also worth trying the ‘dhtlib-update’ branch.
"It always takes longer than you expect, even when you take into account Hofstadter's Law.", Hofstadter's Law
Update:
no success with 4.7k nor with 100nF cap (extreme values again during last 12 hours) - will try the mentioned newer Homeduino sketch and back soon…
Update again:
after upgrading to the newer Homeduino sketch DHT-22 seems to work till now (will see after 12 hours or so)
BUT:
–> every 433Mhz controlled RF device is not manually switchable via webinterface anymore indeed results in “oops - timeout” popup
strange thing: rules are still working ?! ideas?
Graceful bump 
@Cybolord said in DHT-22 sensor outages in use with Homeduino:
strange thing: rules are still working ?! ideas?
This sounds odd. Please restart the system and try again
"It always takes longer than you expect, even when you take into account Hofstadter's Law.", Hofstadter's Law
@mwittig said in DHT-22 sensor outages in use with Homeduino:
This sounds odd. Please restart the system and try again
did this several times meanwhile - old sketch build restored and all works like before, but DHT-22 extreme values are back so I guess the filter plugin would be the best choice indeed - thank you all
@Cybolord said in DHT-22 sensor outages in use with Homeduino:
did this several times meanwhile - old sketch build restored and all works like before, but DHT-22 extreme values are back
ok, I need to revisit the homeduino sketch and have a closer look what is causing the issue. This will take some time, however
"It always takes longer than you expect, even when you take into account Hofstadter's Law.", Hofstadter's Law
@mwittig said in DHT-22 sensor outages in use with Homeduino:
ok, I need to revisit the homeduino sketch and have a closer look what is causing the issue. This will take some time, however
great, thank you in advance Marcus!
btw: there should be some other guys with the new sketch running on their Arduinos, aren’t they?
possibly they have same or none issues at all?!
