I personally have an SCD30 monitor which claims the same accuracy and like it. The important thing is to calibrate them in outside air to 400ppm.
[0] https://www.digikey.com/en/products/detail/senseair/006-0-00...
[1] https://github.com/Senseair-AB/Sunrise-Examples---Arduino
So far people that were interested in the combination of enclosure and pcb or the whole kit.
Does the sensor in your project need to be taken outside to calibrate?
I stopped when I realised many sensors (like scd30/40) need this making it very impractical.. Has to be done once while powered on the whole time in a 7 day period.
https://sensirion.com/products/catalog/SCD41/
https://www.digikey.com/en/products/detail/sensirion-ag/SCD4...
Personally I've been very happy with Sensirion's sensors and I've played with the SCD41 quite a bit. The trouble with it for hobbyists is the assembly. It has what seems to me to be pretty picky requirements for soldering. I've attempted to do it myself but I keep managing to pop off the metal covering.
Haven't tried the Senseair Sunrise but such low power requirements are very very appealing because they mean the sensor can be battery-powered. The SCD41 isn't bad but powering it for a year means a lot more batteries. With a 1 min sampling period, the SCD41 uses 1.5mA while the Sunrise uses 21µA.
I think they're around 60 bucks (e.g. the T6793) so they aren't the cheapest around, but they're good quality.
They have a normal header you can solder a few wires into. Probably the best approach for the Sensirion sensor is to oversize the pads a bit and use paste to re-flow underneath.
The way I soldered the SCD41 is with a pre-made PCB footprint, some solder paste and a mini PCB heater. I think the problem was too high heat. I need to try it again with low-temperature solder paste.
But for my application the Senseair Sunrise looks a lot better.
As a diver I use oxygen sensors, and those should be calibrated once every third day, and linearity can go bad from everything between 6 months to a few years.
If I went with a DIY solution i would definitively go with 3 sensors so one can at least use voting to determine which sensors to trust, and to measure relative drift.
This can still situation where either a bad sensor is marked as good, or a good sensor marked as bad, so a human operator is still required to monitor the decisions made.
The outside ppm changes quite a bit in cities. CO2 dome effects can push it 2x higher.
https://www.researchgate.net/figure/The-maps-of-global-histo...
Older study https://www.sciencedirect.com/science/article/abs/pii/S13522...
... tldr a higher planetary boundary layer over cities can form under various weather conditions. That reduces mixing, and cities are large sources of CO2.
So I can say - I stand corrected, and TIL :)
Anywhere that traps smog should show similar correlations with CO2.
I could even send levels to my home IOT InfluxDB for historical charting.
most people can't, or more precise they probably can put something together but can't make sure it works correctly under any situation (in operation range)
what you pay for is that someone makes sure it works properly
else you can get sensors with similar or same hardware on amazon too, it's just that their correct reliability is not guaranteed at all and just checking if the values outside seem right is by far not enough to verify that (because it e.g. might behave right on low but not high CO2 levels or not on high/low humidity, etc.)
I have seen tests with sensing devices which seemingly where good outside but then e.g. scaled wrongly being way of at higher CO2 levels (either way higher or way lower).
Perhaps you've left the auto calibration on? That's usually a recipe for disaster (and one that can be avoided by reading the datasheet).
This includes all the sensors integrated in these off the shelf products like the author is using, which could explain the extremely high readings.