homeclimate V1.5
Since I have a NAS/server running a TIG stack (Telegraf/InfluxDB/Grafana) for monitoring, it is straight forward to log all sensor data to this machine instead of the Raspberry Pi. The server is consuming a low amount of power, it is running 24/7 to offer its services around the clock and can deal with a little bit of additionally data no problem. This also has the benefit to deliver long time series much quicker than the small Raspberry Pi. Plotting half a year of 30s intervall data in Grafana on a Raspberry Pi 3B+ is a nice stress test – it maxes out the CPU for several minutes and puts a lot of stress on the SD-Card. Apropos SD cards: A reminder to a note in the first part of this buildNever, and I mean NEVER, run influxdb or any other database for more than a short test on an SD…
home climate V1
In 2018, a colleague mentioned to me that he found a nice CO2 monitor with an undocumented but open USB connection. On Github, vfilimonov wrote a nice python package to decode the USB connection and show the measured data on a website. From this starting point on, I experimented with further sensors and a more complex data acquisition, storage and visualisation. The current state is an experimental setup running on a Raspberry Pi 3B+. Several sensors are connected to test and compare them. Connected sensors are: TFA Dostmann AirCO2ntrol Mini CO2 Monitor Two DHT22 temperature and humidity MH-Z19 IR absorption CO2 sensor 0-5000ppm TSL 2561 light intensity sensor BMP280 digital barometer The monitoring system in the configuration as used in the following. Raspbian Buster runs from an old SSD that also stores programs and databases (black cable with two USB port because the Pi cannot supply enough current). The white…