CO₂ sensor for Home Assistant ​

A client contacted me to automate his mushroom production, for which he requires:

• Temperature control
• Humidity control
• CO₂ control
• Remote monitoring and parameters adjustment

Since he didn't have a central automation system yet, I suggested and designed around Home Assistant, which offers tons of functionalities and works from the local network on an old laptop or a Raspberry Pi. It's also easy to add more devices to it, and it integrates well with popular ecosystems like Amazon Alexa and Google Nest.

Hardware ​

We went for the simplest, fastest, easiest design, which includes:

• Adafruit's Feather ESP32 V2 development board
• Adafruit's SCD-30 CO₂, temperature and humidity sensor
• A small, ventilated enclosure for the sensor
• A JST SH 4-pin cable (also called STEMMA QT or Qwiic, or whatever)
• A power strip with smart plugs (in this case, TP-Link's Kasa HS103 or KP303 were good choices for our goal)
• A Raspberry Pi 4 to host the Home Assistant instance

When it came to sensor choice, given the small differences in price, we went for the most precise and simplest, all-in-one solution. Other contenders would have been the SCD-40, SCD-41, and MH-Z19 sensors.

With the hardware choice settled, I placed the order and parts arrived 2 days later.

Software ​

I went with ESPHome to run the sensor with Home Assistant. Tasmota is another great firmware option, though while it offers more settings and flexibility, I find it more complex and less user-friendly, especially for this kind of simple project.

The first time programming the Feather must be done via USB, and the easiest way is probably to use the Home Assistant ESPHome add-on. However, I've never had any luck with web serial, so I went with the command-line setup method instead.

In any case, after running the setup wizard and before uploading the configuration file to the Feather, I modified co2-sensor.yaml like so:

esphome:
  name: co2-sensor01
  platformio_options:
    board_build.f_cpu: 80000000L    # Slow down ESP32 to reduce power.

esp32:
  board: adafruit_feather_esp32_v2
  framework:
    type: arduino

# Enable logging
logger:

# Enable Home Assistant API
api:
  password: "api_password"

ota:
  password: "ota_password"

wifi:
  ssid: "Super-Duper-Wi-Fi"
  password: "SupErS3cretP@$$w0rd"

  # Enable fallback hotspot (captive portal) in case wifi connection fails
  ap:
    ssid: "Co2-Sensor01 Fallback Hotspot"
    password: "SupErS3cretP@$$w0rd2"

captive_portal:

# New section

# Optional: Enable Deep Sleep for spaced out measurements
# deep_sleep:
#   run_duration: 10s
#   sleep_duration: 10min

i2c:
  sda: 22
  scl: 20
  scan: true
  id: i2c_bus_1

sensor:
  # https://esphome.io/components/sensor/scd30.html?highlight=scd30
  - platform: scd30
    co2:
      name: "Mushrooms01 CO2"
      accuracy_decimals: 1
    temperature:
      name: "Mushrooms01 Temperature"
      accuracy_decimals: 2
      filters:
        offset: -0.2   # To be calibrated
    humidity:
      name: "Mushrooms01 Humidity"
      accuracy_decimals: 1
    address: 0x61
    update_interval: 10s

esphome:
  name: co2-sensor01
  platformio_options:
    board_build.f_cpu: 80000000L    # Slow down ESP32 to reduce power.

esp32:
  board: adafruit_feather_esp32_v2
  framework:
    type: arduino

# Enable logging
logger:

# Enable Home Assistant API
api:
  password: "api_password"

ota:
  password: "ota_password"

wifi:
  ssid: "Super-Duper-Wi-Fi"
  password: "SupErS3cretP@$$w0rd"

  # Enable fallback hotspot (captive portal) in case wifi connection fails
  ap:
    ssid: "Co2-Sensor01 Fallback Hotspot"
    password: "SupErS3cretP@$$w0rd2"

captive_portal:

# New section

# Optional: Enable Deep Sleep for spaced out measurements
# deep_sleep:
#   run_duration: 10s
#   sleep_duration: 10min

i2c:
  sda: 22
  scl: 20
  scan: true
  id: i2c_bus_1

sensor:
  # https://esphome.io/components/sensor/scd30.html?highlight=scd30
  - platform: scd30
    co2:
      name: "Mushrooms01 CO2"
      accuracy_decimals: 1
    temperature:
      name: "Mushrooms01 Temperature"
      accuracy_decimals: 2
      filters:
        offset: -0.2   # To be calibrated
    humidity:
      name: "Mushrooms01 Humidity"
      accuracy_decimals: 1
    address: 0x61
    update_interval: 10s

If using the command-line, the configuration file can be added to the ESPHome Add-On by creating a new device, and then editing the configuration file there with what was uploaded to the Feather. The add-on will then detect the sensor as if it was setup via Home Assistant.

Finally, I added the Node-RED add-on to create my automation rules, the TP-Link Kasa Smart integration to control my smart plugs, and the ESPHome integration (not the add-on) to access the new CO₂, temperature and humidity sensor readings. A simple History graph card on the dashboard then lets me inspect the sensor readings.

A steady airflow around the sensor is required to get accurate readings. Contact me at hello@connaxio.com for help with calibration, control loop, and installation tips.