Bosch Sensortec Community

    Showing results for 
    Search instead for 
    Did you mean: 

    BM688 Heater Profile: Are they important?

    BM688 Heater Profile: Are they important?


    Recently I tried out the BME688, using the Bosch AI Software. For my understanding, the heater profiles do the follwing:
    The sensor heats up to a certain temperature, stays there for a certain time and takes the measurement. Then it heats up/cools down to next defined temperature and so forth.

    What is not clear to me yet, is the correlation between temperature and resistance for a certain mixture of gases FOR REPEATED MEARSUREMENTS.
    I know, that different compounds react differently at different temperatures with the surface of the sensor, but why would I for example want to measure the same temperature twice:

    200°C  250°C 350°C 250 °C

    Wouldn't it be sufficient to measure every temperature once  (and maybe vary the measuring time)? In the predefined heater profile I could also find some profiles with repeated measurements at low temperatures and some very short peaks with high temperatures. Are those peaks supposed to "burn" free the sensor?


    9 REPLIES 9

    Community Moderator
    Community Moderator

    Hello hhammer,

    It means to detect different kind of gas, it needs to do much lab test and generate profile.

    I suggest that if you design a project, you could refer current existed profile.

    I understand that the gas resistance differs for different sensor temperatures (e.g. 220°C or 350°C), but this is no explaination why the order in which I measure at differente temperatures would affect the gas resistance.

    Lets say I want to find H2S:

    I measure gas resistance at 250°C, afterwards at 280°C and again at 330°C, each time I measure for 100ms. As a result I get 3 gas resistance values corresponding to the three temperatures, but in my understanding they are not connect in any way to each other. So it would not have any influence if I measure at 250, 280, 330 °C or  330, 250, 280 °C, except for I have to wait until the sensor has reachted its temperature.

    The meaning if the heater profile is therefore unclear to me and it seems like nobody can really explain what the order of temperatures would matter for the measured gas resistance. 

    Any specific hints are welcome!


    Any ideas?

    Community Moderator
    Community Moderator

    The resistance characteristics of different gases are different, and the corresponding profiles are different. Usually, the detection of a certain gas needs to be strictly tested in the laboratory. Otherwise, if the heating temperature is arbitrarily modified, the gas cannot be detected correctly.

    You can use current existing profile.

    Occasional Visitor

    Hi BSTRobin, like you I'm interested in looking at gas mixtures in ambient air. When i worked I spent my time looking at the effect of different gases on both metal and organic semiconductor sensors. Normally it will take severa seconds to get anything like a stable reading, looking at a response after say 140 msecs will give some information about the sensor resistance as it heats up, it will not be an equilibrium gas reading.  Looking at mixtures will complicate this as different gases will adsorb and react at different rates, so the sensor resistance will be the sum of the different gases adsorbed at that temperature.

    I've just bought the BME688 system in the hope of using multiple sensors at different temperatures to try to look at gas mixtures. I've worked with sensors set at 3 temps, which it is claimed detect different gases because of different sensitivities with temp, but looking at normalised data from large data sets I'm coming to the conclusion that tin oxide sensors are being affected more by humidity than gas effects, correlation of gas effects with temp, press and humidity show a marked correlation between humidity and gas effects.

    So I'm not sure whether or not this approach will be any better, but if I can find a way to programme the temp profiles I'll do what you appear to be suggesting, I don't hold out a lot of hope of success, but it will be interested when I can sort out how to program the temp and  duration of a test. Hope you have some luck!