|
Post by rogers1313 on Jan 24, 2019 11:35:39 GMT -8
Since the very first of my project for the Grove Weather Pi, I have had glitchy issues with the AM2315 Humidity Temp probe. I have noticed that several other people seem to have the same issues. When I worked on this early it was that that whatever is wrong with it was related to the OLED display that would not work. Well after updating everything to 3.11 the OLED works, the AM2315 is still glitchy. I reported this issue within the first month of owning this system. My GroveWeatherPi.py file is the latest version 3.14 not 3.11. How do you debug this device? And is there a fix for the glitchy stuff going on. I have the history set for one hour. the outside temp is all over the place from good actual temp to very very large deviations. this station is setup in my house do conditions are very constant right now. Can this be noise related on the I2c network? Attachments:
|
|
|
Post by rogers1313 on Jan 25, 2019 14:26:16 GMT -8
I am going to mount all my components into a new enclosure. Try to organize the wiring to cause a little more separation. Things are a jumbled with my layout right now. are there any tips for grounding on the I2c signal wires with projects like these. Proper signal grounding on other things I have done in the past has helped when things get glitchy. Any ideas out there? Thanks again. Roger
Ps I have looked at recent issues in the AM2315 section of the forum. It looks like others are having issues too.
|
|
|
Post by rogers1313 on Jan 25, 2019 14:51:49 GMT -8
|
|
|
Post by SDL on Jan 27, 2019 16:22:22 GMT -8
Roger,
Hmm. I read the article. That is an interesting comment. A delay between the temperature and humidity reading. I'll check it out.
BP
|
|
|
Post by rogers1313 on Jan 28, 2019 7:42:13 GMT -8
So do you think that the data that is being delivered from the probe may not have enough time to get to the raspberry pi at times and this may result in the resulting CRC error? Things just don't add up at certain times.
My other thought was could this probe be moved to BUS 3 and be handled totally individually.
I am just beginning my next phase of the weather station which is to layout a note how all this software works from the top to the bottom.
Roger
|
|
|
Post by SDL on Jan 29, 2019 9:00:40 GMT -8
Well, that's not quite it, but close. The article puts a delay between reads of the AM2315 when it is awake. That might be something I had not tried. It isn't a matter of the time it takes to get to the Raspberry Pi, it seems to me that maybe two quick reads to the AM2315 hoses up the chip inside of the AM2315. I'm looking forward to trying this next week. I'll issue a new version of GWP so everybody can try it. I've got two things to add to the AM2315 driver. One from Peter and one from the above article.
BP
|
|
|
Post by rogers1313 on Jan 30, 2019 11:00:56 GMT -8
I went to adafruit.com and looked through their notes on the AM2315. Please refer to this link learn.adafruit.com/am2315-encased-i2c-temperature-humidity-sensor/python-circuitpythonThey have a warning note on the raspberry pi section about this probe and the raspberry pi. "At this time this sensor does not appear to work well on the Raspberry Pi. We’ve seen mixed success, but YMMV."If this is true, then there must be a bug with this product and the raspberry pi. Does SwitchDoc have a resolution for a substitute probe that will work better?I hope that your insights from Peter and myself will help resolve the issue. It is the only reason I am not deploying my Weather Station, yet. thanks again for all your help. Roger
|
|
|
Post by SDL on Jan 31, 2019 11:42:01 GMT -8
Roger,
I think we are very close. I think we have much better success in general than the warning above. And circuit python adds a whole other layer of places where than can be issued.
We haven't given up on the AM2315 yet. I do have a couple of other probes coming slow boat so I'll post those results here.
BP
|
|
|
Post by rogers1313 on Feb 15, 2019 15:33:30 GMT -8
I noticed you released version 3.15 of GroveWeatherPI.py. So we can copy that and download to our weather station and give it a try? Are any other files in the software package updated with that that need to be downloaded?
|
|
|
Post by SDL on Feb 16, 2019 11:37:46 GMT -8
Yes definitely. Do deploy. Recent feedback from Peter says that it didn't seem to do him any good. One more change to try before we have to dig in again.
Watch for 3.16 next week.
We are close, but not quite there. I have a solution that works all the time (involves shutting off the power to the AM2315 when it hangs), but that requires a Grove PowerSave.
BP
|
|
|
Post by rogers1313 on Feb 16, 2019 12:45:14 GMT -8
I just rebuilt the software install from ground up following procedures found in several places on the SwitchDoc website and the SwitchDoc GitHub readme file. I am still basing my install on the NOOBS 3.0.0 with the automatic updates that happen when the install is done that way. I found a section that suggest running the following: git clone github.com/adafruit/Adafruit_Python_GPIO.gitcd Adafruit_Python_GPIO sudo python setup.py install Before my rebuild from scratch I had done the first two lines only and did not do the sudo python setup.py install I decided to go ahead after I got the system running and the glitchy am2315 probe stuff that was going on to try this setup. It produced about 10 lines of errors and would cause the software to crash on the weather station. So I rebuilt from scratch and eliminated the 3 install lines above from my rebuild procedure. Guess what the AM2315 problem will run on version 3.15 for 15 minutes before glitching out. This is just trying different things. If you would like I will send you my whole rebuild procedure. Do I need to install the first two lines above for any reason?
Am looking forward to version 3.16 I have added an Blynk one hour history for the output temp. It has longer durations but still is bottoming out. Attachments:
|
|
|
Post by SDL on Feb 16, 2019 14:35:35 GMT -8
Keep it coming. We will figure this out with more data like yours.
BP
|
|
|
Post by rogers1313 on Mar 29, 2019 9:34:31 GMT -8
Any news on the humidity probe fixes?
|
|
|
Post by rogers1313 on Mar 29, 2019 10:53:14 GMT -8
I just found the support thread for the AM2315. It has some great updated information.
|
|
|
Post by SDL on Mar 30, 2019 13:05:58 GMT -8
Woohoo! And boy, our latest driver (not released yet on GWP) seems to be rocking. It should go up there this weekend.
BP
|
|