I was curious about how precise the "local" weather forecasts are and ended up ordering a cheap weather station "kit" from Ali Express so I could compare my measurements with the measurements that come from various "hyper local" weather forecast services.
When deployed, it's supposed to look something like this:
Each of the sensors plugs into a controller board that looks like this:
Eventually, a box of sensors and a controller board arrived in the mail. There was no included documentation and what little additional documentation that I have found online since ordering and deploying the station has been collected and added to the docs folder. Small tweaks to the code included in this repo based on what i've learned from the documentation.
Each of the sensors is quite simple, but w/o any documentation, it's impossible to know how much rain fall each pulse from the rain fall sensor equates to... etc.
I didn't want to spend much time trying to reverse engineer the sensors or figure out how to drive them, so I spent most of my time trying to work with the bundled controller. I did get some decipherable documentation via google translate and, after a bit of trial and error, managed to get a PoC Arduino sketch that could decode the packets. After a bit more work, the Arduino code was ported over to the ESPHome platform. That's what's included here.
C/C++ is not my preferred language, so the code "works" but could probably be optimized in quite a few ways. Pull Requests are accepted!
A more detailed writeup and a basic 3d printable enclosure for outdoor deployments is written up here.
In short, the WS3 supports a specific particulate matter sensor (Note: Affiliate link; thanks for your support!). Making a cable to interface the sensor with the WS3 module isn't too much work and is discussed in the detailed writeup linked above.
I've updated the custom_component for the WS3 station to support the sensor. Just set the SUPPORT_PM25_SENSOR to 1 and run esphome your_file.yaml run --upload-port ... to build and upload a new FW version.
While hacking on the additional support, these existing repos were helpful in figuring out which values from the PM2.5 the WS3 was forwarding:
- https://github.com/rpanfili/airQualityMeter
- https://github.com/i3water/Blinker_PMSX003ST/blob/master/src/BLINKER_PMSX003ST.cpp
The datasheet has been included for reference.