BME680 bVOC value very off

[mhoff87]

I have been experimenting with the BME680's gas sensing using 5 BME680 breakout boards inside a small gas chamber. I am using an ESP32, BSEC2 library, and sampling every 3 seconds on LP mode. Here is one test I ran a few times:

- Ran the BME680's for multiple days until IAQ accuracy reached 3

- Flush chamber with ultra zero air until readings stabilize (https://www.calgasdirect.com/gasco-105l-1-zero-air-calibration-gas-20-9-o2-balance-nitrogen-a-105-li...)

- Fill chamber with 1ppm isobutylene in Nitrogen calibrated gas until readings stabilize (https://www.calgasdirect.com/isobutylene-calibration-gas-c4h8-1-ppm-balance-nitrogen-in-a-103-liter-...)

The BME680's bVOC values immediately get pinned at 1000ppm when the gas enters the chamber. I understand it is calibrated to a different gas mixture, but there isn't anything that can explain a 1000x difference. Isobutylene has a molecular weight of 56.1 g/mol and I calculated the bVOC gas mixture to have a molecular weight of 51.29 g/mol. I would appreciate any help or thoughts people have.

Thanks!

[mhoff87]

Bump

[BSTRobin]

Hi mhoff87,

Isobutene, as a volatile organic compound, can theoretically be detected by BME. I have some questions here:
(1) Is the second injection of gas a mixture of isobutene and nitrogen, with isobutene at 1ppm? Is it possible to directly charge 1ppm of isobutene, considering that nitrogen may react with isobutene under certain conditions to produce substances such as isobutyraldehyde and acetone.
(2) Do you consider inflation pressure? High pressure gas can directly flush into the BME sensor pores, which can also have a certain impact.
(3) Are all 5 sensors showing abnormalities? When displayed at 1000ppm, does it continue to display this value? Is there a situation where it drops to normal.

[mhoff87]

1 - Isobutylene in nitrogen is a pretty common calibration gas for VOC probes. I don't have the capability of injecting 1 ppm of isobutylene, I think that would require pure isobutylene and injecting a specific volume? I refered to this document for a lot of my testing: https://tsi.com/getmedia/cff9a1be-02fd-4ad5-b437-1ea664b9a8f7/TSI-156_Field_Testing_Calibration_VOC_...

2 - This seems like a good explanation. The chamber we are using has a check valve to let gas out but not let air in. This could be creating a slight positive pressure. It definitely isnt "High" but it is positive. I have included a picture of the setup.

3 - All 5 sensors are showing abnormalities. They continue to display 1000 until the chamber is opened. I have attached a plot showing the bvoc reading from all 5 BME680s while Isobutylene is charged into the chamber. I have also attached a plot of gas resistance during this same test.

[BSTRobin]

Hi mhoff87,

Could you measure VOC following this order 'air->Isobutene->air', in the meanwhile, it's better to measure pressure, humidity and temperature parallelly. The sensor can not operate in low oxygen environment, you can add some air in the box.
And please provide all the graphs after your test, thanks.