Selecting a combination temp/humidity sensor?

A little background on my project, I’m working on a solution to monitor the temperature and relative humidity in all the rooms of my home and be able to monitor that info from my phone while at home.
I have something now that used DHT11s connected to some ESP-01s and sending their data to an MQTT broker that I can then use to feed a dashboard I can monitor from my phone.

The temperatures I see from the DHT11s are off by enough from some commercial thermometers I have in my test rooms that I’m looking for alternatives.

I have tried looking at a DHT22 but the one I picked up to test does not work well with the ESP-01. I can get readings at initial power on, but a reset or wake from deep sleep results in my not being able to get data from the sensor until it is power cycled again. My guess is that noise on the data line on reset puts the sensor in a bad state. So I’m looking at other options

So I’m trying to identify some other sensors to look into. I’ll be looking to hopefully get 8-10 sensors once I’ve settled on one. That means I’m looking for sensors that don’t cost a lot.

Here is what I am trying to look into now. Almost all of these seem to be available for $1.50-$2.00 when imported from China unless otherwise noted.

AM2320 - Since it appears to be basically a DHT22 but with an I2S interface, it seems like it would be OK. But after running into issues with the DHT22 I’m approaching this one cautiously.
It looks like they are about $1.50-ish if ordered from China which isn’t the cheapest of the bunch, but isn’t that expensive either.

BME280 - well thought of but pricey at about $10 (the ‘cheap’ ones seem to be mislabeled BMP280s). I’ve only included it here because of the respect people appear to have for this sensor.

AHT20 - I don’t see much on what people think of this sensor outside of being able to use it on Arduino. I’m going to lump in the AHT10 and AHT15 with this since the specs seem similar and there is support, but there is even less info on what people think of these 2 sensors.

Si7021 - I’ve seen it used in a few projects but I don’t see much about the pros and cons of it.

SHT30 - I don’t really know much about this one other than there are libraries for it.

HDC1080 - I only saw this one while trying to dig deeper into the topic of available sensors so I have no idea about this one.

Are there any gotchas or other information about these senors that should make me consider one over another?

check this out: Compare DHT22, AM2302, AM2320, AM2321, SHT71, HTU21D, Si7021, BME280

1 Like

If your error is consistent considering using an offset to correct the value. The technical spec states: DHT11 20-90%RH at 0-50 ℃ accurcy 5%RH, temperature ±2℃ ( ±3.6F). From what you state it appears it is working properly but I do not know anything about the accuracy of the commercial thermometers or anything about them being in valid calibration.

1 Like

The relative humidity is off by enough with the DHT11s to not be usable. The temps aren’t too bad once the sensors have time to ‘settle’.
I’d like to do a ice water/boiling water calibration with them, but between not being waterproof and the limited range of the devices, I can’t really do that.
The RH inaccuracy is a totally different story. It’s a very important number to determining the comfort level of a room. I don’t think I can calibrate the DHT11’s issues away there.

The problem I’m having with DHT22s is they are finicky things. I’m using ESP-01s modules to run the sensors and report the data to an MQTT broker. I’m using deep sleep to save power. The DHT22s will read fine at power on, but will give me NaNs after the controller wakes from deep sleep or is reset. I guess noise on the data line puts the device in a bad state.
In a separate test, it seems more in line with what I was seeing with other thermometers and seemed to better represent the ‘room comfort level’. But its quirks are part of what sent me here looking for alternatives.

Good data to have!
A conversation elsewhere suggested the SHT30 which is similar enough to the SHT31 that it looks like it’ll do fine. And I think it is likely to behave better when paired with an ESP-01s than the DHT22s I was testing out as ‘drop in’ replacements for my DHT11s (but they are far too finicky to be usable).