Bosch Sensortec Community

    Showing results forΒ 
    Search instead forΒ 
    Did you mean:Β 

    Ozone exposure damaged BME680?

    Ozone exposure damaged BME680?

    New Poster

    I have a couple BME680s, both of which were exposed to ozone in the range of 0.2ppm to 0.3ppm. One had maybe 5 hrs exposure, the other, maybe 1 hr total.

    This seems well below the permissible limit stated in the BME680 layout considerations document:

    It is not permitted to expose high concentrations of reducing (e.g. >20 mg/m 3 of TVOC) and/or oxidizing gases (>10
    mg/m3 of NO x +SO x +O 3 ) to operating sensors for a long time

    Doing some calculations, 10mg/m3 of O3 is almost 5ppm.

    Please see attached graphs of static AIQ output from BSEC. "ozone (aiq)" had less exposure than "filter (aiq)". Each point on the graph is based on 20 samples taken over 60s averaged down to a single sample. The sensors shot way up and have been very slowly working their way back down. They both have been in fresh air for almost 12 hrs now.

    Going from memory, the raw resistance reading of the one that saw less exposure is about 20-30% higher than it previously was in fresh air.

    So, are these fried? If so, lesson learned πŸ™‚

    BTW, running BSEC 1.5.1474 on ESP32 via Arduino library. Polling at BSEC_SAMPLE_RATE_LP.

    3 REPLIES 3

    Community Moderator
    Community Moderator

    @tofu wrote:

    So, are these fried? If so, lesson learned πŸ™‚

    If the sensors are still reactive, within gas resistance range, and on the trend to stable values, most likely they are not fried πŸ˜‰

    Ozone is an oxidizing gas, thus based on the exposed concentrations it will increase the gas resistance of the sensor. Outdoor air is quite different in compounds and concentrations than typ. indoor air, and BSEC's IAQ outputs are optimized for indoor use-cases.

    Thanks @handytech.

    The IAQ readings seemed to remain at an abnormally high level. Maybe I just wasn't patient enough. I tested both sensors by waving around a cotton swab soaked in isopropyl and they were both still very responsive.

    I ended up just power cycling the sensors and clearing the stored BSEC state from the ESP32's non-volatile storage. IAQ readings seem to be back to normal. See attached graphs.

    Technically this was "indoor" air, although it wasn't breathable :). I guess it's just that the BSEC algorithm isn't tailored to heavily contaminated air? Or was there actually a permanent change to the resistance of the sensors?

    Community Moderator
    Community Moderator

    In this case BSEC (and MOx sensors in general) is not so much affected by the concentration ("heavily") of the pollution, as much as to which compounds are present and their effect on the raw sensor signal.

    An offset in the gas resistance value is not a damage to the sensor nor directly affecting its sensitivity, but the BSEC library would take some time to self-calibrate. Clearing BSEC's state and restarting the library as you have done also does the job.