BSEC config for 2.8V

[The_DudE]

Hello,

BSEC library comes with few config "blobs" for both 1.8V and 3.3V, but nothing in between. Thats a shame as I guess quite a lot of "single-supply" IOT devices using primary batteries + DCDC will stand right in the middle to optimize battery life while keeping the rest of the system correctly powered (1.8v is still not mainstream for radio tranceivers and sensors).

Is there a way to request additional configs from BOSCH (like 2.5V that would stand in the middle or 2.8V)?

What is the impact of using "wrong" config voltage ?

Regards,

Laurent

[o_o]

With BME680 in LP mode, the part heat up enough to have an impact on the temperature measurement.This is modeled into the compensated temperature output.

Since the parts draws constant current, the higher the voltage, the more power is dissipated by the internal linear voltage regulator

Therefore if you use a config string for the "wrong" voltage :
Raw temperature output is good
Pressure output is good
Humidity output is good
IAQ outputs are good.

Compensated temperature output will either over, or under compensate for self-heating.

[The_DudE]

Hello,

thanks for your reply,

If I understand well, as I'm planning to use the sensor in ULP mode, the impact of using "wrong" config should be marginal, as the 5mn pace should be enough to let the chip go back to ambient temperature?

Other question regarding the config 4Days/28Days flavor: is this just the time-window accounted for min/max values or is it also impacting dynamic response (filters)?

Best regards,

Laurent

[o_o · Accepted solution]

---

@The_DudE wrote:

If I understand well, as I'm planning to use the sensor in ULP mode, the impact of using "wrong" config should be marginal, as the 5mn pace should be enough to let the chip go back to ambient temperature?

---

Indeed in ULP mode, the sensor has plenty of time to cool down to ambiant (on your PCB) temperature. The part itself will not cause heating. Temperature measumrement is always measured before gas measurement takes place.

Regarding the 4Days / 28Days options. This is the time constant of the filters that track the history of the sensor. The way to think about it is : 63% of the data older than 4 (or 28) days has been forgotten. Here are the recommendations:

Mobile device -> 4 Days

Stationary device -> 28 Days