the SCD40 sensor is a huge upgrade from the MH-Z19D. it's a dual sensor photoaccoustic NDIR design that uses one as a calibration reference to avoid drift.
I'm still including the MH-Z19D anyway, since I happen to have the sensors hanging around, and it'll be interesting to compare drift.
the other major upgrade this time is in the power delivery. I've added a buck-boost to ensure a stable 5V supply instead of just using whatever VUSB is (which is rarely actually 5V), to avoid sensor skew.
for the non-5V stuff I'm also doing all my own sensor voltage regulation onboard for the sensors, using LDOs with known PSRR specs, rather than relying on the ESP8266 devboard's onboard 3.3V LDO like I did last time.
I've also since learned my lesson on proper capacitor selection, so there's not a single 100nF cap in this entire design. the majority of the decoupling is 1uF 0402 and/or 10uF 0603.
@gsuberland oh yeah, it’s just the “exactly 1” that threw me. Usually for any given size/value/rating combination of passives there’s either zero because they are impossible to make, or quite a lot
@gsuberland how do you calibrate a scd40? It uses the min value over 24h iirc, but can you tell it the actual co2 contents corresponding to this min value ?
Add comment