<_whitenotifier-4>
[glasgow] electroniceel commented on issue #220: Qualification tests for revC2: pre INA233 - https://git.io/JICRb
Stormwind_mobile has quit [Ping timeout: 246 seconds]
Stormwind_mobile has joined #glasgow
<electronic_eel>
more progress: i just read the first current values from the INA233 on my revC2 with some hacked firmware. values seem to be reasonable
Stormwind_mobile has quit [Read error: Connection reset by peer]
Stormwind_mobile has joined #glasgow
<d1b2>
<0x53A> With the SCD30, how drastic would the effect of a wrong calibration be? Is it relative or absolute? If I vent for a few minutes, co2 goes down to ~500ppm. Outside is apparently supposed to be ~400ppm (though I doubt I can exchange 100% of air in that short timeframe). Assuming it is miscalibrated to 500 when its actually 400, would that mean it is wrong by 100ppm, or by 25%?
<d1b2>
<Attie> have you checked the datasheet / specs?
<d1b2>
<Attie> ... it claims +/- 30ppm +3% MV (though I'm not currently sure what that is) - repeatability is also +/- 10 ppm
<d1b2>
<Attie> i.e: if the environment is 400ppm, I'd expect a min reading of 370ppm, and a max of 430ppm
<d1b2>
<Attie> there is also the drift to consider - there's a note in the datasheet about exposing the sensor to a 400ppm environment "regularly" to keep the +/-50ppm drift over lifetime in line. it also suggests that environments with <400ppm can result in drift.
<d1b2>
<Attie> ah, looks like MV = "measured value"
<d1b2>
<Attie> ... i i guess the max reading for a 400ppm environment could be as much as ~442.9ppm (+30 +3%)
<d1b2>
<0x53A> I read that it just remembers the lowest value it ever encountered over a 7 day timeframe, then assumes that is 400ppm and calibrates everything else against that. So theoretically, if the lowest value it ever encountered was 1k ppm because you never vented, it would be offset by 600ppm.
<d1b2>
<0x53A> Ah, I read the datasheet again, and it says Accuracy Drift over lifetime: +-100 ppm
<whitequark>
you have to enable auto calibration for that 7 day thing to kick in
<d1b2>
<Attie> interesting - I read this to mean that after activating ASC, it requires "fresh air" (i interpret that as ~400ppm, though "fresh" can be less than that) for >1hr per day, over a 7 day period... it must remain powered for the full 7 day calibration cycle
<d1b2>
<Attie> ... the command permits enable/disable of "continuous" ASC, and it's off by default
<whitequark>
yep
<d1b2>
<Attie> also, re +/-100ppm... the sheet(s) i have here state +/-50ppm, which is interesting (v0.92 and v1.0)
<d1b2>
<0x53A> ah yes, with a search I found some "preliminary" datasheet, the official one says +-50
<d1b2>
<Attie> ah, oky
<d1b2>
<0x53A> ok, so if I never activated auto-calibration, it should still be factory calibrated with a maximum offset of 50 ppm or a bit more
<d1b2>
<Attie> there's also the FRC (forced recalibration)... i can't see for sure, but I'd guess (??) that they do this in the factory as part of it's calibration
Stormwind_mobile has quit [Ping timeout: 240 seconds]
<whitequark>
you can also just do it yourself
<d1b2>
<0x53A> you mean put it outside and run recalibration?
<whitequark>
yes
<d1b2>
<0x53A> I currently have it on my desktop computer, and the usb cable isn't long enough to get to the window, I'll see if I can just move the computer. But I was hoping it was just massively miscalibrated, and that seems unlikely. I just woke up, started the computer, and saw a reading of 3.6k. Now I am concerned 😐. And the math checks out, unfortunately. I have 40 m^3, the "rise" rate of the graph is
<d1b2>
<0x53A> ... the "rise" rate of the graph is ~ +400 pmm per hour, which would amount to 0.78 kg of co2/day, websites say a human typically exhales a kg, so the magnitude is correct
<d1b2>
<0x53A> though 400 ppm/hour * 8 hours means I'll always wake up at +3200 ppm