The Adafruit_BME280_Example.py code outputs the sensor.t_fine value with a label of "Timestamp", but from what I can tell this is actually a "fine temperature" value used in pressure and/or humidity compensation calculations. It doesn't appear to have anything to do with time and is probably of little value to anyone except in debugging. See for reference section 4.2.3 of the datasheet http://www.adafruit.com/datasheets/BST-BME280_DS001-10.pdf .

I think it would be nice to have this as a PyPI package. Is anyone looking into this already?

Dear Friends,
I am doing a master project in IOT!
How can I Publish and subscribe RPI 3 -BME280 sensor data in MQTT broker using Python script??
I know the simple message subscription and publication python programs, but to publish BME 280 sensor data and subscribe the published data in MQTT protocol, I don't know. Could anyone help me with this issue?
Thanks in advance!

Regards,
Thangaraj

From what I can understand of the datasheet (6.2.3 Compensation formulas), it seems that there is a typo in the code: the value 16384.8 is used instead of 16384.0.

It produces a little error in the humidity value.

read_raw_temp() blocks for the conversion-in-progress indication to drop before trying to read. While this may be appropriate for the BME280 "forced" mode, data is available at all times for "normal" mode, as far as I can discern from the data sheet and my experience.

In the case that high-rate operation is desired, BME280_STANDBY_0p5 reduces the standby time between conversions to 0.5 ms. As a result, the use of time.sleep(0.002) will often "miss" the window between when the conversion is completed and another begins. What should be jitter of the conversion period itself, now can be several periods. In early testing, I was seeing delays that often were 100-300 ms with a conversion period of under 50 ms.

If anything, the check should probably be done on bit 0, the "im_update" flag. It probably should be done with a block read of the sensor data, as I imagine that the duration of the time that the data in the registers many not be valid is very short. As that is described as an "NVM" data copy, I'm not sure that it even impacts the sensor registers (see data sheet section 7.4.4).

Should the library ever be expanded to support "forced" mode directly (the ability to write the proper registers is not exposed in the current API) there would be a need to have a blocking read until the conversion is complete. Until that time, this blocking behavior is detrimental for high-rate sensing applications.

For anyone reading this, one work-around is subclass BME280 and override read_raw_temp() removing the blocking wait

        # while (self._device.readU8(BME280_REGISTER_STATUS) & 0x08):    # Wait for conversion to complete (TODO : add timeout)
        #     time.sleep(0.002)

Hi,

when running it on a Pi Zero with a BMP280 connected via i2c it returns 0.00 for humidity:

root@zero:~/Adafruit_Python_BME280# python Adafruit_BME280_Example.py 
Timestamp = 114731.000
Temp      = 22.409 deg C
Pressure  = 986.51 hPa
Humidity  = 0.00 %

Is there a way to get the humidity?

Hi,

I modified the library to be able to put the device in sleep mode (almost no power consumption) when I don't need it and to support a soft-reset, because sometimes my sensor gets stuck and reports the same temperature for hours. Until today I always just unplugged all wires when that happened. Next time I'll try to reset the sensor.
I would like to contribute these changes here (if you're interested, that is), but I have no clue how github works. I need a pull request, right? Do you have a little tutorial for me to start with? :)

Greetings!

I believe the fourth line of the package install instructions ("cd Adafruit_Python_GPIO") should read "cd Adafruit_Python_BME280".

Should be

SDA0/1 to the SDI pin