search

Ribbit-Network / ribbit-network-frog-hardware

The sensor for the world's largest crowdsourced network of open-source, low-cost, GHG Gas Detection Sensors.
https://www.ribbitnetwork.org/
MIT License
94 stars 26 forks source link

SCD30 CO2 Sensor High Humidity Bug #17

Closed keenanjohnson closed 3 years ago

keenanjohnson commented 3 years ago

Something odd is going on with one of my test CO2 sensors. About 2 days ago, it started reporting a relative humidity that was radically high and hasn't yet returned to normal. The blue line below is higher than real humidity:

image

I pulled the sensor out and let it dry out inside for a few hours, but that didn't change anything.

Then I tried to bake it at around 80 C for a few hours in the oven. That did seem to have improved the humidity sensing, but it seems like I fried something in the CO2 LaserElectric torchFireRightwards arrowFrowning face. 1.2K PPM is NOT correct.

image

SCD30 Serial Number: 2862279

keenanjohnson commented 3 years ago

I'm not too concerned about the sensor failing after the oven testing, I must have just exceeded some thermal limit, but the humidity sensor reading incorrectly seemingly randomly seems concerning.

I'm going to reach out to Sensirion to see if they have any insight on why that would happen.

keenanjohnson commented 3 years ago

@dahlwinters have you ever seen anything like this?

keenanjohnson commented 3 years ago

Email out to the Sensirion: image

dahlwinters commented 3 years ago

Haven't seen anything like this with the relative humidity. If it was with the CO2 readings only, it might have been an auto-calibration problem. Definitely something is wrong since a value of RH above 100% is not possible.

keenanjohnson commented 3 years ago

Thanks! I'll update this issue as I learn more!

keenanjohnson commented 3 years ago

More info from Sensirion, but no root cause.

"Thank you for the addition temperature plot. The graph shows a steady temperature drop of around 10°C and at the same time an unreal humidity increase over 100% RH. The temperature decrease is nearly linear and looks basically the same as if someone entered a temperature offset. The SHT31 on the SCD30 PCB is limited to an output of 100% RH. The only way to get values over 100% RH is setting a temperature offset as the compensation engine output is not limited to 100% RH. The average temperature offset due to self heating of the SCD30 PCB is 3°C. I’d suggest to set the offset to 3°C and see what happens.

There are a few other points to consider.

keenanjohnson commented 3 years ago

Closing this issue out for now. I doubt I'll root cause the sensor issue without more help from the vendor, but I've documented what I know here for the future in case we see this issue again.