High CO2 readings on Sen66 Outside

ESPHome
,
1

Hi there,
I’ve got a Sensor for Air Quality that I built around one year ago, and it used to live in the bathroom, now it’s got a new case and lives outside (sheltered from rain ofc). What’s confusing me is the fact that it shows 700ppm of CO2, when it should be half of that. The PM sensors are also way off, I don’t believe the readings are that high (especially when compared to freely available air data from not that far away). Is the sensor cooked, and I need a new one?
Thank you in advance :slight_smile:

grafik
grafik416×461 17.2 KB

The esphome sketch uses the now native Sen6x profile (copied everything below the usual backup hotspot, etc.

captive_portal:
  

web_server:
  port: 80

fan:
  - platform: speed
    output: casefan
    name: "Case Fan"
    speed_count: 100
    
    
output:                                                                                                                                                                           
  - platform: ledc
    id: casefan
    pin: GPIO2 #D1                                                                                                                                                                            
    frequency: "25000 Hz"                                                                                                                                                                                                                                      
    min_power: 30%
    max_power: 100%

time:
  - platform: homeassistant
    id: homeassistant_time
    timezone: "Europe/Berlin"

i2c:
  sda: GPIO5 #D4
  scl: GPIO43 #D6
  scan: true
  id: bus_a

sensor:
  - platform: sen6x
    type: sen66
    id: sen66
    pm_1_0:
      name: PM <1µm Weight concentration
    pm_2_5:
      name: PM <2.5µm Weight concentration
    pm_4_0:
      name: PM <4µm Weight concentration
    pm_10_0:
      name: PM <10µm Weight concentration
    temperature:
      name: Temperature
      id: sen66_temp
    humidity:
      name: Humidity
      id: sen66_humidity
    voc:
      name: VOC
    nox:
      name: NOx
    co2:
      name: Carbon Dioxide
    update_interval: 10s
2

Unlike the senseair CO2 sensor I don’t think you can force a manual calibration after leaving it outside where the CO2 level should be ~420ppm.

It is supposed to happen automatically for the sen66.

3

PM 1 is really high also, is sensor located next to some combustion? That could explain both.
What about humidity, is 31% realistic in your location?

4

Yeah I was looking at the manual and everything, seems like there’s no good way to autoclean or recalibrate it. Maybe I’ll write to Senserion to ask about it :neutral_face:

5

No, there’s nothing that should be giving off any sort of dust apart from plants that give off pollen; Again, that doesn’t excuse the high numbers either, because I have another one of those sensors in the kitchen not too far from this one, and even with all the windows open, the PM never goes that high (except for cooking and smoke from that for example).

30% humidity is right on target, weather service and air gradient sensors not too far away also show the same readings: around 400 ppm CO2, around 3um PM2.5, etc.

grafik
grafik430×263 14.1 KB

6

PM1 is not dust but smoke.
Take the sensor back inside and try if it settles.
Also, SEN has internal fan and sensirion provides pretty detailed instructions for enclosure design. Yours have some external “case fan” also?

7

I’ve read Senserion’s guides for enclosures, yes. The sen66 isn’t the only thing in the box outside, there are a bunch of other things, and it’s powered by poe (since it lives right next to an outdoor switch). Things got quite hot in the past so now there’s a fan to pump the hot air out. Without that, the box gets hot, and the sen66 reads the same air over and over. Easily proven by breathing into the box and seeing how long it takes for the CO2 and such to settle back down to normal.

All the other stuff is “after” the sen66 so the temp doesn’t get biased by the components. I don’t think I’m even close to hitting the high limit for airflow around the sensor since it’s a 40mm noctua fan at around 40%, and the intake is much bigger.

I’ve thought about it for a bit and I think I’ll just buy a new sensor. PM and CO2 haven’t changed at all, even with wind and multiple resets to get auto clean going (in this case powering the sensor off and on), so I’m assuming that the PM sensor must be covered in dust or something. Maybe I accidentally cooked the sensor.

8

Sen6x internal airflow design is quite sophisticated to keep the optics clean.
Anyway, before you throw it in trash been, disconnect it from other hardware and give it a try alone. Observe esphome logs to see any misbehavior.
Frozen i2c bus for example..