Question about overall calibration and why my numbers are all over the map?

Components
#1

So after getting everything put together and Sketch uploaded the values coming back seem very odd.

Temp 24.70 while 2 separate other thermostats read arount 21-22 Celcius
Humidity on sensor reads 38 vs 45 on separate unit.
PM2 has read 0 since I’ve installed it.
CO2 is all over the place Right now it’s reading 65343? which I think I would be dead if that was the case.

Everything was soldered properly, and I uploaded the sketch that Jeff Geerling had on his post, so I’m wondering what’s going on.

Realistic values for PM 2.5?
#2

My PM2 was always reading 0 initially as well, so I resoldered the wires to the holes in the board and then started getting low numbers. After moving the unit outside, I was getting 5-8 and Weather service was showing outdoor numbers were 7 that day, so I guess it was just that clean inside.

Your CO2 jumping around certainly isn’t looking right

#3

@Nomad
This is what you can try:
a) Ensure you use a high-quality power adapter. The CO2 sensor seems to need a stable power supply.
b) Temperature sensor inside the case heats up from the other elements. Look at our new enclosure with the external temperature probe to get higher accuracy.
c) To test the PM sensor, create some smoke and see if it reacts.

Also please post pictures of your built (back and front side) and we can have a look.

1 Like
#4

OK so I ended up re-soldering everything with some additional flux to get the solder in the holes, and it seems to be better.
Temps are the same, but close enough to the standalone thermometers I have.
PM2 now reads 6
CO2 now reads 97?? so that seems low, I’ve tried to jump the pins a few times where it brings the value up to around 800 or so, but it eventually drops down low again.

Do I just keep retrying to jump the pins or is it expected to eventually self calibrate?

Photos
image

image
image714×487 36.8 KB

Backside
image

#5

I am not sure if it will revert back to the 400ppm calibration automatically.

I did not find anything in the spec sheet.

On page 5 they talk about manual calibration.

#6

Following this thread, as I about did the manual calibration as described by AirGradient, but now worried about messing something up, so leaving it be for now.

#7

Normally it takes anywhere from 3-10 days for the sensor to self-calibrate and after that, it should work very stable.

#8

** update, so after trying to reset the CO2 a couple more times I seem to be getting better numbers now, still a bit low 322, but it does go up and above 400 from time to time, and if you breath on it then it does spike… so at this point I’ll give it time to auto calibrate.

The PM2 also reacts to something like a candle burning nearby with appropriate values. The only thing left to tweak is the temperature. For that I’ve opened a separate: Question about correction factor in the code for temperatures

Cheers.