CO and NO2 readings unrealistically high and static

[silvijovince]

Hi,

I am currently using the Apollo AIR-1 with the optional MiCS-4514 gas sensor.

This is already my second AIR-1 unit, as I previously returned the first one due to the same issue.

I am observing consistently elevated values:

CO ≈ 4 ppm
NO2 ≈ 0.15–0.17 ppm

The problem is that these values appear unrealistically high and remain almost completely static, even under conditions where I would expect near-zero readings.

To rule out environmental factors, I have:

• placed the device outdoors (balcony, open air) for extended periods
• ensured there are no local pollution sources (non-smoking environment, no combustion sources)

Despite this, the readings remain essentially unchanged.

At the same time:

• SEN55 NOx index is very low (typically 2–8)
• VOC index is also low (~50)

This suggests that the air is clean, which contradicts the reported CO and NO2 values.

Additional notes:

• Device is running factory firmware (no ESPHome modifications)
• The behavior is identical across two separate units

Questions:

1. Are CO/NO2/NH3 values from the MiCS-4514 expected to be only indicative rather than actual ppm concentrations?
2. Is it normal for these values to remain nearly constant regardless of environmental changes?
3. Is there any required burn-in or calibration procedure for the MiCS-4514 that could explain this behavior?

At this point, it is difficult to interpret these readings as meaningful, so I would appreciate clarification on expected behavior and limitations of the gas sensor module.

Thanks!

[bharvey88]

Thanks for opening this issue we appreciate it! I have already replied in email and discord but I will follow up here as well in case any others see this in the future!

This is normal and started happening once we submitted a fix to esphome to make the component esphome uses for the mics to match the datasheet. esphome/esphome#7173

It's possible the esphome mics component could be further improved, however we already have it matching the datasheet from dfrobot (the makers of the mics gas sensor) so we are not sure what else could be changed. You can open an issue on the esphome repo to see if they can help triage if you'd like!

Closing as non-issue and further issues should be submitted directly to esphome repo for MiCS gas component