Go Down

Topic: BME680 (Read 1 time) previous topic - next topic

ElCaron

Does anybody have experience with this senors? I like using the BME280, and I was pretty thrilled when I learned about the BME680, but looking into it deeper, a few things sound strange to me.
- For once, since it uses MOX, it doesn't sound as low power as it is advertised.
- I don't really understand the power modes. Ultra-low power mode seems to consume more than 6 times more energy than low power mode. Is that for cases where my supply cannot even provide 0.9mA? It seems best to disable the gas sensor most of the time and only fire it up occationally to make a reading in 1.4 seconds.
- Does that even work at least halfway accurately? Adafruit recommends to have it running for half an hour before reading, which pretty much prohibits all long term battery application.
- Can we expect meaningful readings from an ambient temperature sensor right next to a 320°C hotplate in a tiny package?

knut_ny

Quote
- Can we expect meaningful readings from an ambient temperature sensor right next to a 320°C hotplate in a tiny package?
datasheet page 19
Ny

jremington

Quote
- Does that even work at least halfway accurately?
See pages 8-9 of the datasheet.

wvmarle

- Can we expect meaningful readings from an ambient temperature sensor right next to a 320°C hotplate in a tiny package?
Make sure your solder doesn't melt.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

ElCaron

See pages 8-9 of the datasheet.
Are pointing me to the overall accuracy specification? I don't see any inforamtion about longer pre-heating phases there, however I DO see some references to "new sensor" in the datasheet, indicating that there are long- and maybe short-term differences.

datasheet page 19
Are you talking about the compensation calculation?

Make sure your solder doesn't melt.
How am I supposed to understand that?

knut_ny

datasheet page 19 points to warmup period (ca. 30 ms) before reading the sensor.

Ny

wvmarle

How am I supposed to understand that?
The melting point of solder is just below 200°C. At 320°C your sensor may fall off the PCB, assuming it can handle the temperature in the first place (it's almost certainly out of spec).
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

ElCaron

datasheet page 19 points to warmup period (ca. 30 ms) before reading the sensor.
I have seen that, but it doesn't tell me if Adafruits info is nonsense or the datasheet is too optimistic.

The melting point of solder is just below 200°C. At 320°C your sensor may fall off the PCB, assuming it can handle the temperature in the first place (it's almost certainly out of spec).
Well, you might want to tell that to the Bosch engineers (page 8 of the datasheet linked above)

wvmarle

BME680 data sheet, page 8:
Quote
Operational range: min -40°C, max 85°C
So 320°C is indeed out of spec, as I expected already.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

ElCaron

BME680 data sheet, page 8:
So 320°C is indeed out of spec, as I expected already.
We are talking about the heated gas sensor. On the page I mentioned:
"Heater target temperature
320 °C, constant operation
(VDD ≤ 1.8 V, 25°C)"

"[Peak Supply Current] Occurs within first ms of
switching on the hotplate"

sego

#10
Nov 25, 2018, 12:54 pm Last Edit: Nov 25, 2018, 01:23 pm by sego
I´ve now started with this sensor, too. It shall replace the BME280. But as far as i can tell, it´s not really usable for battery configurations. It´s the first sensor where i have to use the Adafruit library, because the others don´t work for me. And even without using the gas sensor, it takes about 1 second to update, which is very much. Even without filtering and oversampling, the readings of the Adafruit library are unbelievably slow. I don´t know why.
The next thing is the gas sensor itself. The standard value in the Adafruit library (setGasHeater(320, 150)) isn´t usable at all if the sensor isn´t constantly read. However, it cannot be constantly read when using it on battery. So you have to play with the values. I am using setGasHeater(380, 1200) at the moment. This is giving me round about the same values when read once an hour. every reading is influencing the next reading. For example: If you start the station at 10:50, the second value measured at 11:00 will be off, because the time span of 10 minutes will influence the reading (higher) when all other readings will be between 60 minutes. I assume that this is due to rest heat in the sensor.
The consumption with arduino is 20mA without the heater for about 1 sec + 40mA with the heater for about 1 sec. So you have 30mA in 2 sec which is a constant 0.8mA in a minute when read once an hour. The rest needs about 0.2mA of constant current. So you see, it´s a massive reduction in battery life. Using 9 AA cells (3 in row, 3x parallel) which all together have about 9000mAh (3x3000mAh @ 4.5V), you drop from 1875 days of battery life to 375 using this sensor once an hour.
I haven´t really understood how to use the ultra low power mode and if this could be an alternative, but the normal mode on battery is, in my opinion, crap. The overall power consumption is simply too high.

Go Up