paulvha / svm30

Sensirion SVM30 on Arduino, ESPxx, 32U4, Lorawan, DUE
10 stars 1 forks source link

SGP30 Baseline issue #3

Open mikewolfd opened 1 year ago

mikewolfd commented 1 year ago

Hello,

When I use your update_baseline script, after it sets the baseline, it remains "locked" and stops changing. Is this expected behavior? Shouldn't the baselines continue updating?

Mike

paulvha commented 1 year ago

Hi Mike long time since I worked on this. In principle yes it can change. Make sure that auto-calibrate is on after setting baseline. That said I see on epshome

The SGP30 sensor will re-calibrate its baseline each time it is powered on. During the first power-up this will take up to 12 hours. Exposing to outside air for at least 10 minutes cumulative time is advised during the calibration period.

But it is calibrated to the best values, I would not expect it to change a lot (although it can). Maybe try to expose the sensor to a different (internal or external) environment

regards, Paul

mikewolfd commented 1 year ago

@paulvha Thanks for responding!

I haven't seen an "auto-calibrate" setting anywhere, I'd love to see information on that?

I've just been seeing that if I start from a raw state, and give it 12 hrs to calibrate, the read baselines will continue changing, although slowly. But if I were to re-instantiate the baseline on reboot from these saved values, I don't see the read baselines update at all anymore, they're frozen.

I see you mentioned it here but the wording is confusing.

paulvha commented 1 year ago

I looked again but I mistakenly called out auto-calibrate as I was working on another project with a sensor that has auto-calibrate function and mixed the open PDF's. Also with respect to the link to the example3, so it seems nearly 3 years ago I had the same issue. Doing 4 to 5 different projects in a month... It has slipped my mind. Still, I never moved the SGP to an outside air environment to give that a try, maybe that works, maybe it only works in the 12 hours.

In the datasheet May 2020 for the SGP30 there was also talk about "Inceptive Baseline for TVOC measurements". Not explained further how it impact really.

background On many of the environmental sensors, there have been open questions. e.g. never seen an explanation about the many calculations for the BOSCH BME280 or BME680? The SPS30 from Sensirion is using an internal (not exposed/explained) algorithm to calculate some of PM results. There are also many sensors that just start to continuously provide a number... But in the many projects I have done, it has been as good as impossible to find 2 different sensors to produce the same result. Then again.. these are relatively cheap sensors (not professional) and they provide a good indication.

regards, Paul