Just got back from 2 weeks of work in Europe so the project is behind a bit more than I would like. Did some more testing today and I have confirmed that the MQ-2 sensor indeed is not perfect for taking this project to the "next level" in commercial applications. It performs great but I can attest that with my code it has an annoying weakness with discriminating between the 3 levels if saturation is high.
With that said, it does perform great as long as you don't need discreet levels. It seems to trigger just fine- only the measurements can be unreliable as it canl announce a gas or level that isn't fully accurate (basically 2 levels will announce but only one is truly in the test environment). I am still trying to source a commercial grade sensor that I can afford to integrate- but for now I think I may just mod the code so it is strictly a go/no go indication of a dangerous condition. For this it seems to work great! No issues at all with even a slight amount of smoke or gas.
I will also try modding the code to attempt better resolution between the gasses. I have my doubts this can be done reliably. A proper commercial grade sensor seems to be the path forward but I am willing to try :)
This is simply the best part of inventing for me- working out minor hurdles and ending up better than the original target.
Next up a video of assembling the unit with my custom PCB's linked on this project. First test seems to work great!
Full assembly video tutorial coming soon! Cheers