Is it me, or is SGP-30 simply useless? : )

Sensors Environmental
1

Hi, I’ve been using SparkFun’s Air Quality Sensor - SGP30 (Qwiic) for over a week now, and have been running dozens of tests, and all of them have totally. I’m either doing something awfully wrong, or this sensor is totally unreliable (which I’m guessing is not the case).

I have set-up a CO2 measurement test chamber (about 1meter X 1meter X 1 meter). In this chamber I test the CO2 ppm levels with an industrial grade measurement device. Outdoors, the (proper) values I get are about 420-500, in our office it’s around 450-600 on a normal day, and inside the chamber it fluctuates between 450-500 on idle, and when we pump CO2 into the chamber goes up to 4000-5000 (and possibly even over it).

However no matter what I do with the SGP-30, where I place it (outdoor/indoor/in-chamber) the values I get out of it (running the basic/raw/humidity examples) are usually around 400-500 (even if I spray lots of CO2 quite close to it).

  • I’ve left it in the chamber for a lot more than the initial 15sec required for calibration.

  • I’ve tried initializing it outdoors and only after a a few minutes bringing it in for the “official” measurements inside the chamber. I’ve also tried initializing it indoors and doing the same, and of course also initializing inside the chamber.

  • I always have another (expensive and reliable) CO2 measurement device next to it.

  • I’ve read all the other related posts on this forum (just a couple though), and I’ve re-read all the docs and guides.

  • I’ve purchased another SGP-30 and have verified its measurements are wrong as well.

  • I’ve added the humidity sensor for relative humidity compensation.

  • I’ve added a fan into the measurement chamber.

  • I’ve verified the chamber is sealed enough and the air inside it is well-circulated.

Please, help us soon. We’re totally stomped. : (

2

wow… looks you have done your homework and have taken all the advice. :slight_smile:

I have worked on an SVM30 (which has the SGP30) and it was reacting as expected. Fast and accurate . Just breathing into the sensor would make it react in output. The humidity compensation hardly made a difference for me in the outcome.

Brainstorming :

On some sensors, I have seen a sticker to prevent dust during transport. Could that be an issue here ? (maybe stupid question… but anyway)

I found the SGP sensitive to the voltage (1.8V). In the case of this board, it should be on the top left pin of the 5 pin regulator (AP2112K) above the connection points

3

Consider studying the SGP30 data sheet closely, and set up your box to measure the ethanol and H2 reference signals, after 20 hours of burn in.

Keep in mind that this is not a “CO2” sensor, it responds to potentially thousands of volatile organic compounds (VOCs), any of which may be present in your environment and adversely affecting response.