I found that there is a pre-wired DHT22, so I thought I would try it to see if I can improve the results by getting the device further away from the hot board.
Anyway when it arrived it wasnt what I was expecting at all since I never found a photo of it next to a standard DHT22. Its like asking your parents to get you lego for xmas and they buy you duplo.
Thats right - its a standard DHT22 inside a giant plastic enclosure.
I just solder wires to the pins on the regular DHT22. I’m definitely going to stay away from the pre-wired version unless I need something to occupy an unnecessary amount of space.
So, I have tried a couple of other external DHT22’s with a smaller size.
This red one :
Looks like it has a 4.7k resistor in it.
This black one :
Looks like it has a 5.1k resistor in it.
The problem I have is I cannot get any accuracy out of these external DHT22’s. I have tried 2 of each of them and all 4 give poor accuracy - particularly regarding humidity (such as 20% higher than any other sensors I have).
I looked through the Piettetech code, but I dont see anywhere to change resistance if that is the reason for the difference. I have a bunch of plain DHT22’s which I have used and I put 10K resistors in those and they are all quite accurate. They all use AOSONG DHT22’s in them, so I expect there is another difference I’m supposed to account for ?
Any ideas if there is a different software configuration I should try ?
The communication with the DHT22 is digital, not analog, so the resistor is just a pull-up resistor. As long as you are meeting timing, the resistor value is not very critical at all.
The protocol to talk to these parts uses short and long pulses to indicate zeros and ones. It is the timing that is critical and that is why the Piettetech interrupt-driven version is much more reliable.
Ok, I am just using the standard Piettetech DHT22 code :
I will just use the giant DHT22 since that has no circuitry in it, so I can choose
my own components for it.