5 ms·
I was curious about the specific equipment they were using and found a list of spec sheets on page 11 of [0]. I'm really happy they decided to use models that a
by Youden 5y ago
I was curious about the specific equipment they were using and found a list of spec sheets on page 11 of [0]. I'm really happy they decided to use models that all have manual fresh-air calibration.
A big problem with DIY projects is that most of the sensors have a built-on, on-by-default auto calibration function, which assumes it will consistently see fresh air for ~1h per day (or other period), which isn't always the case in an environment that doesn't have commercial HVAC (like a house or small school).
When the assumption is broken, the calibration is too and the data are close to useless as absolute measurements (they still have some small value as relative measurements).
The assumption that a fresh air baseline will always be seen is particularly easy to break right now. It's winter, so you're less likely to open the windows and due to COVID, you may be working from home, leaving very little time for CO2 to disperse.
[0]: https://www.coschools.org.uk/resources/RP24.4%20How%20to%20Use%20CO%E2%82%82%20monitors%20in%20education%20and%20childcare%20settings.pdf https://www.coschools.org.uk/resources/RP24.4%20How%20to%20U...
- ahaucnx 5y agoYes. It is important to use CO2 sensors that do not have a short automatic baseline calibration like the MH-Z19 that is popular in DIY projects but calibrates every 24h. Better use a sensor like the Senseair S8 that only calibrates every 8 days and you most likely have a weekend of low occupancy that should bring CO2 down to ambient levels.
- 5e92cb50239222b 5y agoI haven't been able to open a window for about a month due to extreme air pollution outside, and I think my Senseair S8 is broken now. If it sees fresh air with normal CO₂ levels it goes from "normal" 1.5-2k PPM (which are really closer to 4k) to zero and then shoots up to ~65000 PPM, and stays there. It spent the last two weeks in a room with an open window and still shows 65k PPM. Does anyone know of any way to recalibrate it, except for the obvious solution of putting it in a closed box and slowly lowering the amount of CO₂ there?
- janekm 5y agoThere is a calibration pin which you can connect to GND for 4-8s to force a manual calibration (if you can bring the sensor into an environment of ~400ppm): https://emariete.com/en/meter-co2-esp8266-nodemcu-sensor-senseair-s8/ https://emariete.com/en/meter-co2-esp8266-nodemcu-sensor-sen...
- ahaucnx 5y agoI think 65k ppm is outside the measurement range of the S8. It might be a communication or code issue that causes this and not from actual measurements.
- nemetroid 5y agoAccording to its datasheet, the S8 reports CO2 level as a 16 bit integer, so values around 2^16 (65535) sound like underflow issues.
- ysleepy 5y agoYou can set the calibration mode to manual using either the serial port or a pin on the mh-z19. Mine is configured to manual and after a few months it still reads reasonable values for outside air. But yes, the default is autocalibration. But honestly, the difference isn't that bad, it's probably like 100ppm offset for stale room air from outside, and 100 doesn't make a huge difference when you are trying to measure around 1500.