Add class attributes to allow more control over the sensor #22

jraber · 2019-02-27T06:52:00Z

Add attributes for IIR filter, standby period, operation mode,
and overscan for temperature, pressure and humidity.

Add Enum classes for the overscan, standby, and iir filter values

I started working on this after reading issue #21 . Admittedly, I haven't tested this (I don't have a BME280), so there may be some silly errors here. If someone is willing to test, I'd bet happy to fix any bugs.

Almost all of these changes could be applied to the BMP280 library with very little effort.

Add attributes for IIR filter, standby period, operation mode, and overscan for temperature, pressure and humidity. Add Enum classes for the overscan, standby, and iir filter values

jraber · 2019-02-27T07:10:30Z

Closing this pull request, as I found a mistake in the overscan_temperature property. I'll submit another PR later (after I figure out why Travis failed)

makermelissa · 2019-02-27T07:40:47Z

Hi, you can just leave the PR open and commit new code to your repo instead of needing to create a new one each time.

jraber · 2019-02-27T08:06:31Z

Oh! Thanks @makermelissa
I'm still getting a fail from Travis CI, I guess because of the pylint message. Should I throw a "disable=too-many-instance-attributes" at it?

Any suggestions are welcome.

makermelissa · 2019-02-27T15:36:16Z

Yeah, that should be fine to add.

makermelissa · 2019-02-28T14:53:40Z

Hi, I connected a BME280 to a Feather M4 Express running the latest version of CircuitPython. However, it is not finding the enum class that you are using and thus doesn't run. I'm curious what your test setup was.

…lter

jraber · 2019-03-01T05:03:56Z

My (limited) testing was with python 3.7.2, but enums has been part of the python standard library since 3.4. I don't have a BME280, so I haven't actually tested with the device. I am very interested to know your results.

I added an examples/bme280_normal_mode.py to show how I expected for it to work. Could you give that a look?

jraber · 2019-03-01T06:50:06Z

So, I'm just realizing the CircuitPython != Python.

If CircuitPython doesn't have enum, I can re-implement those enums using the const type from micropython like is used for the other constants in this library.

makermelissa · 2019-03-01T07:03:31Z

Hi @jraber. Yeah, it doesn't appear CircuitPython currently supports Enums. Also, thanks for updating the example to show off your changes. I'm excited to test your changes.

Removed references to the enum class, while keeping the same functionality Updated the 'normal mode' example

jraber · 2019-03-01T14:19:59Z

Hi @makermelissa . I removed the references to enums. Please let me know if you see any other issues.

makermelissa · 2019-03-01T19:12:40Z

Hi, I went ahead and tested it and the normal_mode example works, but the numbers appear to be off.

Current Code Results

Temperature: 24.5 C
Humidity: 28.6 %
Pressure: 1008.9 hPa
Altitude = 36.16 meters

Changed Code Results

Temperature: 188.8 C
Humidity: 100.0 %
Pressure: 300.0 hPa
Altitude = 9165.37 meters

Also, simpletest doesn't appear to be working anymore. I took a closer look at the code and it appears to now default to sleep mode. Can we change it so that it basically starts the same as before by default, but maybe have the option to put it in sleep mode? I'm sure this library is in use by someone and I don't want them to update and have their code be suddenly broken. Thanks for all your work on this.

0x800000 is the *default* value and is a valid value

Changed_BME280_OVERSCANS from frozenset to dict to allow lookup of the associated value for the measurement time calculation

jraber · 2019-03-07T09:52:14Z

Hi @makermelissa . Thanks for testing! I'm not sure why you're getting odd values from the normal_mode example. I pushed a couple of changes today. If you have a moment, could you test with those changes?

I adapted these changes to the BMP280 library and was able to successfully run both of the examples on a Raspberry PI with the BMP280 connected via I2C.

About the simple example, what did it do when you ran it? Did it not return any values? Crash?

I was trying to keep the default functionality the same as the current version.

You're right that my code defaults to SLEEP mode, but that's actually the default mode for the sensor anyway. When _read_temperature is called, it will change the mode to FORCE (only if the mode isn't set to NORMAL). After each forced measurement is completed, the sensor sets itself back to sleep mode automatically.

makermelissa · 2019-03-08T06:25:43Z

Hi @jraber, I'm not sure what's going on with the code. I tried it with your latest update and I get the following when running bme280_normal_mode.py:

Traceback (most recent call last):
  File "code.py", line 14, in <module>
  File "/lib/adafruit_bme280.py", line 453, in __init__
  File "/lib/adafruit_bme280.py", line 120, in __init__
  File "/lib/adafruit_bme280.py", line 432, in _read_byte
  File "/lib/adafruit_bme280.py", line 461, in _read_register
  File "/lib/adafruit_bme280.py", line 457, in _read_register
  File "/lib/adafruit_bus_device/i2c_device.py", line 115, in write
OSError: [Errno 19] Unsupported operation

If it helps, I've been testing this on the Adafruit nRF52840 Express.

jraber · 2019-03-08T14:39:57Z

@makermelissa , I don't think any of my changes would cause that issue. Especially not the changes that I pushed recently. Could you check the wiring to the sensor? Could you try with an old version of the library, to be sure that the sensors is connected and working correctly?

jerryneedell · 2019-03-08T14:58:16Z

FYI - I just tried running the normal mode example on a Raspberry Pi and it executes normally

makermelissa · 2019-03-08T21:05:04Z

Thanks @jraber. I'll give it another try when I have a moment, unless you want to try it in CircuitPython @jerryneedell.

jerryneedell · 2019-03-08T21:07:26Z

Sorry, I can't test it easily since my bme280 is in a "deployed" raspberry pi. It was easy to test it there, butI don't have one I can test with CP right now.

makermelissa · 2019-03-12T18:07:10Z

I should be able to give this PR some more attention tonight.

makermelissa · 2019-03-13T03:07:16Z

Numbers are still way off for me:

Temperature: 188.8 C
Humidity: 100.0 %
Pressure: 300.0 hPa
Altitude = 9165.37 meters

vs

Temperature: 21.0 C
Humidity: 33.7 %
Pressure: 1006.2 hPa
Altitude = 59.02 meters

I'm going to dig in deeper to see if I can narrow down as to why they're off by so much.

makermelissa · 2019-03-13T03:41:29Z

This one is a head-scratcher. There's nothing obvious about the changes that should make it so the readings are off by so much. I'm thinking it may have to do with some register setting, as that is the major thing that changed. I'm going to continue messing around with this though.

makermelissa · 2019-03-13T04:15:32Z

Ok, I figured out where it was getting stuck on the simpletest at least. It seems to not be getting the status on line 143. I think this is part of some bigger issue and I'm continuing to trace it down. I suspect the numbers are off because in normal mode, it isn't waiting to let the conversion complete since lines 143 and 144 are now being skipped, however that may only apply to forced mode.

In FORCE mode, the sensor changes back to SLEEP after completeing a single measurement, but we are not updating the mode internally. It's better to always write the mode to the sensor, and trust the caller not to change it needlessly.

makermelissa · 2019-03-13T04:58:22Z

I'm getting a status of 255 no matter what. I also forgot to mention I've been using SPI. I'm going to try this with I2C to see if that makes a difference.

jraber · 2019-03-13T04:59:02Z

I suspect the numbers are off because in normal mode, it isn't waiting to let the conversion complete since lines 143 and 144 are now being skipped, however that may only apply to forced mode.

My understanding of normal mode is that you don't have to wait for the conversion to finish; there will always be data to read, either the new measurement (if the conversion is complete) or the previous measurement (if the conversion is still running).

I think the mode setter may have been causing an issue with the simpletest. Basically it would only allow the sensor to go into FORCE mode once. The latest commit may fix that.

makermelissa · 2019-03-13T05:02:21Z

Ok, trying it in I2C mode did make a difference. It works fine in I2C mode, but the numbers are way off with SPI.

makermelissa · 2019-03-13T05:07:40Z

FYI, this library is really finicky in SPI mode (even before this PR) and was much more pleasant to test with I2C.

jraber · 2019-03-13T05:12:18Z

I just (yesterday!) copied the SPI interface from this library to fill a gap in the BME680 library. Go figure, the BME680 gives crazy readings too.

makermelissa · 2019-03-13T05:17:33Z

Yeah, at least we've narrowed the problem down.

makermelissa · 2019-03-13T05:23:56Z

I wonder if it's related to the SPI mode. That would explain why I got 255 for status instead of 0 and why the numbers are so far off. The datasheet says it supports SPI mode 0 and 3.

makermelissa · 2019-03-13T05:32:29Z

Nope, didn't make a difference. It worked the same in either mode.

jraber · 2019-03-13T05:33:06Z

The issue may be that I am enabling SPI 3-wire mode. Bit 0 (spi3w_en) in register 0xF5 (config). I don't think that was set before. Let's turn that off...

makermelissa · 2019-03-13T05:35:03Z

Bingo! That was it.

makermelissa · 2019-03-13T05:36:31Z

I'm gonna try pulling your code again (free of my edits) and test. If it passes, I'll approve.

makermelissa · 2019-03-13T05:42:54Z

Ok, the normal mode example is fine (in fact, all finickyness is gone), but simpletest isn't running in SPI mode. I get:
RuntimeError: Failed to find BME280! Chip ID 0xff
Normally resetting the chip/board would take care of that. Let me see if there's just something that needs to be added/changed. It's so close.

makermelissa · 2019-03-13T05:46:28Z

My bad. I forgot I moved the CS wire. It's working great.

jraber · 2019-03-13T05:53:24Z

Sweet! Thank you @makermelissa for all of your debugging this!

makermelissa · 2019-03-13T05:54:08Z

You're welcome. Thank you for doing this and all of your patience.

Updating https://github.com/adafruit/Adafruit_CircuitPython_BME280 to 2.3.0 from 2.2.2: > Merge pull request adafruit/Adafruit_CircuitPython_BME280#22 from jraber/master

Add class attributes to allow more control over the sensor

3ebd5c3

Add attributes for IIR filter, standby period, operation mode, and overscan for temperature, pressure and humidity. Add Enum classes for the overscan, standby, and iir filter values

jraber closed this Feb 27, 2019

jraber added 2 commits February 27, 2019 01:36

Fix-up overscan_tempearure as it was returning the wrong value

dc89e0b

Increase underline length to make sphinx happy

032d38c

makermelissa reopened this Feb 27, 2019

jraber added 3 commits February 27, 2019 08:35

Default normal_flag to False to avoid UnboundLocalError in _write_config

29af22e

Fix typo: standy -> standby

e9d7320

Fix __write_config, docstring comes first

3284ed4

Add disable=too-many-instance-attributes to keep pylint happy

3b37d55

Added and example for setting the mode, overscan, standby, and iir_fi…

ccf5363

…lter

Fix-up example

a8c12e7

Refactor to remove dependency on the enum class

ef0eef6

Removed references to the enum class, while keeping the same functionality Updated the 'normal mode' example

kattni mentioned this pull request Mar 4, 2019

CircuitPython Libraries! adafruit/circuitpython#1246

Open

jraber added 2 commits March 7, 2019 02:04

Do not return None when temp, pressure, or humidity value = 0x80000

79d4e85

0x800000 is the *default* value and is a valid value

Add properties for the typical and maximum measurement time

f40fb2c

Changed_BME280_OVERSCANS from frozenset to dict to allow lookup of the associated value for the measurement time calculation

Always write the new mode to the sensor

9c37a71

In FORCE mode, the sensor changes back to SLEEP after completeing a single measurement, but we are not updating the mode internally. It's better to always write the mode to the sensor, and trust the caller not to change it needlessly.

Don't enable SPI 3-wire interface

0a10d86

makermelissa approved these changes Mar 13, 2019

View reviewed changes

makermelissa merged commit febcd51 into adafruit:master Mar 13, 2019

makermelissa mentioned this pull request Mar 13, 2019

needs oversampling + IIR Filter options #21

Closed

jraber mentioned this pull request Apr 30, 2019

Adopt changes from the BME280 library adafruit/Adafruit_CircuitPython_BMP280#13

Merged

Add class attributes to allow more control over the sensor #22

Add class attributes to allow more control over the sensor #22

Uh oh!

Conversation

jraber commented Feb 27, 2019

Uh oh!

jraber commented Feb 27, 2019

Uh oh!

makermelissa commented Feb 27, 2019

Uh oh!

jraber commented Feb 27, 2019

Uh oh!

makermelissa commented Feb 27, 2019

Uh oh!

makermelissa commented Feb 28, 2019

Uh oh!

jraber commented Mar 1, 2019

Uh oh!

jraber commented Mar 1, 2019

Uh oh!

makermelissa commented Mar 1, 2019

Uh oh!

jraber commented Mar 1, 2019

Uh oh!

makermelissa commented Mar 1, 2019

Uh oh!

jraber commented Mar 7, 2019

Uh oh!

makermelissa commented Mar 8, 2019

Uh oh!

jraber commented Mar 8, 2019

Uh oh!

jerryneedell commented Mar 8, 2019

Uh oh!

makermelissa commented Mar 8, 2019

Uh oh!

jerryneedell commented Mar 8, 2019

Uh oh!

makermelissa commented Mar 12, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

jraber commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

jraber commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

jraber commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

jraber commented Mar 13, 2019

Uh oh!

makermelissa commented Mar 13, 2019

Uh oh!

Uh oh!

makermelissa commented Mar 13, 2019 •

edited

Loading

makermelissa commented Mar 13, 2019 •

edited

Loading