Is Your CO2 Sensor Really a CO2 Sensor?

Lesson learned: a metal oxide (MOX) sensor cannot detect CO2. If you read the thick, boring part of the data sheet they go in to detail about calculating “eCO2”.

It kind-of makes sense and it’s also kind-of bogus. Before we could reliably detect volatile organic compounds we could detect CO2. So we
used CO2 as a sort of index into how fresh air is with the assumption that “fresh air” removes VOCs and other pollutants from inside an air
space. Which, arguably it does, but if you’re bringing in “fresh” air using only the definition of “lower CO2” you’re completely ignoring
actual PM (particulate matter) and tVOC rates.

So, MOX sensors are cheap and they can detect tVOC.  If you look at the data sheet for the ENS160 from ScioSense it goes in to the
detail of having four MOX sensors and what they can detect (documentation on this sensor).  I think this going to be my sensor upgrade for tVOC.

Here’s the key phrase from their data sheet:

“The ENS160 reverses the proportional correlation of VOCs and CO2, by providing a standardized output signal in ppmCO2-equivalents from measured VOCs plus hydrogen, thereby adhering to today’s CO2 standards, as shown in Figure 3.”

So, in fact, it can’t detect CO2. If I had a lot of CO2 released in to the air, up to dangerous levels? No warning from this sensor.

I talked to a high school friend who is a retired chemist, she said that a good CO2 sensor (like they would use in a lab) is not cheap.
And she’s right! The one we use in ROCIS, Green Eye CO2, runs around EU 250. There are cheaper ones on amazon but I don’t recognize the brands.

Sparkfun has this Sensirion chip that does claim to detect CO2 but the initial calibration process takes a week