BE VERY CAREFUL when looking for a device which measures CO2 or CO
DO NOT use any kind of air quality sensor as a substitute for a dedicated and certified carbon monoxide monitor. The consequences may be lethal
Many “multi-in-1” air sensors are only useful as a way of indicating that a window should be opened/extractor fan switched on (eg: when the kitty litter box has just been used)
If you do want to measure CO2, look for a device containing a NDIR sensor
The problem:
Many/most of the cheaper sensors out there are simply using TVOC (alcohol) sensors and inferring all the other gases (including HCHO, CO and CO2)
The CO2 readings these things give is “EQUIVALENT CO2” - which is not an actual carbon dioxide measurement
Equivalent CO2 is the atmospheric warming effect of any given gas compared to actual CO2 - eg, methane is 100-20x as potent as CO2 (1), so its eCO2 would be equivalent to a methane reading multiplied BY 20-100
The CO is “Equivalent CO” and based on assumed combustion products for the volatiles detected
In other words - these meters are not measuring actual gas levels, despite their claims
HCHO/eCO/eCO2 levels should be discarded if you’re using them in air quality calculations - and TVOC reports should be taken with a tablespoon of salt (just use them as an indication of poor ventilation)
The really dangerous part:
TVOC based CO readings give both “false positives” and “false negatives”
NONE of the TVOC-based sensors I’ve tested show any sensitivity to carbon dioxide or carbon monoxide - even when lethal levels of CO are tested.
On the flipside they all give pegged readings of CO/CO2 if exposed to butane, propane or alcohol(2) with reduced sensitivity to methane(3)
What this means is that cursory testing can result in a false sense of security because the device is displaying a CO level that appears to change over time
NB:
Whilst it’s possible to measure CO with a TVOC detector it requires a “heater off” method that I’ve never actually seen used (it’s not reliable in any case, being extremely temperature dependent, requires individual device calibration AND regular recalibration)
Apart from the obvious step of testing the devices in a closed box, detecting these “fake” units is fairly simple:
If you graph the reported gas levels you’ll find all the claimed “pollutants” will track each other virtually identically - which is a dead giveaway
Examples of fake meters are the commonly available 2C08 “Tuya X-in-1” unit sold from a number of sources and JQ300 sensors (search for these terms and you’ll find them I’ve purchased and torn down a few). The PM2.5 sensors in these units aren’t great either - being designated by their maker as being for indicative use only (this can be remedied with individual calibration but it’s really not worthwhile)
Actual (and accurate) CO2 meters are devices like the PV28 family (PV28-CW has a decent PM2.5 sensor too)
Good luck and don’t get caught out
(1) Methane has 100x the warming potential of CO2 in the scale of 10 years, 20x at 100 years
(2) Alcohol sensors are cheap and have broad sensitivity to hydrocarbons
(3) If you want to detect natural gas (methane), use a dedicated methane sensor. By the time a standard TVOC sensor is alarming, you’re already got explosive mixtures on your hands