That project goal is to mesure CO2 and if too high, tell user to open their windows. It’s a simple way to remind to recycle air and so, reduce risk against COVID-19 in crowded space (like classrooms). Inspiration is from Eole project. Thanks
Optionaly, you can send data to a server (influxdb & grafana) to get values over time.
- esp8266 (I’ve used a Wemos D1 for tryout, and then an ESP8266 for definitive project)
- I2C 128X64 OLED LCD screen
- Humidity & temperature DHT11 sensor (needed as input to SGP30)
- SGP30 Air Quality Sensor CO2 sensor
- GND on GND
- VCC on 3.3V
- SCL on esp8266 D1
- SDA on esp8266 D2
- GND on GND
- VCC on 3.3V
- SCL on esp8266 D1
- SDA on esp8266 D2
- GND on GND
- VCC on 3.3V
- sensor to d5 / GPIO14
First copy parameters.h.dist to parameters.h.
It contains the CO2_THRESHOLD value which inverts the display.
It’s a simple way to bring attention to users to open windows.
As explained on Adafruit SGP30 sensor page, the SPG30 sensor needs to be calibrated with a baseline value (given by the sensor). That baseline is periodically asked from SPG30 sensor and stored into EEPROM (~1 time per hour). When the system boots, it's read from EEPROM and the SPG30 is calibrated with that value. That increase the sensor precision.
If I understood correctly, the best baseline value is get after 12 continuous hours of operation. If you don’t use the sensor for 7 days, you’ll have to let it run again for 12 hours to « reset » the correct baseline.
Also, the DHT11 sensor is used to compute absolute humidity level, also used to increase SGP30 precision.
If you define a WIFI_SSID, a WIFI_PASSWORD and a INFLUXDB_URL in parameters.h, the esp8266 will periodically send temperature, humidity & co2 to an influxdb database. You can set the location name (like the room where the sensor is placed) in LOCATION parameter.
You can install influxdb, telegraf (used for system metrics) & grafana via docker (links in reference).
You’ll have to create a sensor datbase via : curl -XPOST -s http://localhost:8086/query\?pretty\=true --data-urlencode "q=CREATE DATABASE sensor"
You can then send a test data :
curl -i -XPOST 'http://localhost:8086/write?db=sensor' --data-binary 'temperature,location=home value=20.0'
And read it back :
curl -XPOST -s http://localhost:8086/query\?pretty\=true --data-urlencode "db=sensor" --data-urlencode "q=SELECT * FROM temperature"
The grafana board I configured has been exported to grafana directory.
Here’s a sample output :
