Ported my AQ meter from Clue, plus AIO - but weird display c

[ryliepavlik]

I've been running the combination of an scd30 true co2 sensor and the Stemma QT pm2.5 sensor for a while on my clue, just because it had the Stemma QT port, a display, and plenty of RAM. The FunHouse was an ideal transition, though, since it's got WiFi so I can make it report to my MQTT server eventually (just doing Adafruit io right now). After a little tinkering, factoring out the Simple Text Display class from the Clue library (which works great on the FunHouse if you don't want all the power of portal base), I've got it going! I am seeing a weird reporting interruption in IO which I've posted about in that forum. However it does basically work, and the pir is much more user friendly than using the proximity sensor to control display on vs sleep (I'm terrified of burning in an image, I think because the ghost you see at display startup once made me think I burned in my TFT gizmo in like an hour of tinkering), though more distracting if you're sitting in front of it (since it will flick on and off).

Sometimes I do see a weird display artifact after a while (a few hours), and I'm not sure why. Not sure if it's memory/framebuffer corruption, signal integrity issue, or what, but it always looks approximately like this, extra yellow pixels (looks like this time it also got an extra zero on the 0.3um line, there's normally nothing to the right of that because there's no pm0.3 measurement). Goes away with a reboot so it's not permanent, just weird. Don't remember ever seeing that with Clue, though I usually didn't let it run overnight since it wasn't reporting out to anything but the screen.

Code is here: it's all either based on permissively licensed examples, or my own permissively licensed work, so feel free to reuse parts or the whole thing. https://gist.github.com/rpavlik/23e359f ... 18c5639b6f