In preparation for my first “real” arduino project — a bathroom lights/vent fan controller using humidity and motion sensors — I’ve whipped up this temporary rig:
Inside the box there’s an Arduino Nano, 16×2 LCD and some glue components. 5V comes in from the black wire on the right. Sticking out the top-left (short wires) there’s a WiFi module and the at the end of the longer wire are two temperature / humidity sensors (blue and white thingies).
The goal of this thing is to capture some data about how the humidity changes in a bathroom. I want to figure out whether the cheaper blue sensor is worth using (spoiler: it’s not), when to turn on the bathroom vent fan and later turn it off, and in general I’m just curious about stuff like how much difference does the fan make, etc.
The rig takes a sensor reading every few seconds and sends it over WiFi to my computer. For a quick test I ran it for a bit this morning, then I dangled the sensors inside that water bottle on the right for awhile, waited for things to stabilize and took it back out again. The test ran for an hour:
Here are the results: solid lines are temperatures (degrees Celsius on the left Y axis) while dotted lines are relative humidity (%, right Y axis). Green lines from the white sensor (a.k.a. the good one) and red lines from the blue sensor (a.k.a. the crap one). At 08h14 I moved the sensors into the bottle and took them out again at 08h45.
The first thing I learn from this is that the cheaper blue DHT11 sensor (plotted in red) is terrible. Within 20 seconds the good one jumps to 80% RH while the crap one takes 3 minutes to rise up 20% from 35% to 55% RH and takes more than 15 minutes to approach a stable value while the good one reaches stability within a minute. Both return to baseline more quickly after being removed but again the crap one takes longer. Interesting that it comes down much faster than it goes up.
The crap one’s precision is +/- 1 degree while the good one is tenths, in case you were wondering why the temperature from the red sensor looks wacky. Both show a slight temperature increase right after being exposed to the high humidity chamber and both show a slight temperature decrease after coming back out (although this is harder to see with only 1 degree precision on the crappy one).
The DHT11 (crappy blue sensor) costs around $1.50 on eBay free shipping from China while the DHT22 (good one) is around $4.50. The Arduino Nano board is around $3.25, the WiFi module is around $3.50. Because the WiFi module runs on 3.3V and will die if it gets the 5V that the Nano runs on, I needed a logic level converter ($1.50) and I used freecycled diodes to drop the 5V supply voltage to around 3.3V.