|
|
Post by SDL on Sept 8, 2018 17:03:57 GMT -8
Excellent. That is the confirmation that I was looking for.
Now we will reproduce this in the laboratory (use The PiDatalogger to hammer the in-house OurWeather system until failure!)
What is the frequency you used to get the 12 hour failures?
I'm thinking we may have to move to an MQTT based interface to get rid of this problem for all time. BUT, there are a number of things I can do once I have a failure mode.
BP
|
|
|
|
Post by edharris on Sept 10, 2018 5:48:32 GMT -8
Update: Second repro (still with 2 minute timing) happened 19 hours, 50 minutes after the power cycle.
|
|
|
|
Post by SDL on Sept 12, 2018 7:26:09 GMT -8
Thank you!
BP
|
|
|
|
Post by triggerfish on Sept 20, 2018 2:49:39 GMT -8
I just see this thread and realise I had a bunch of NaN values myself yesterday on GroveWeatherPi... Are driver updates "global" on github or isolated by project? Will I have to update stuff on my GroveWeatherPi also? |
|
|
|
Post by SDL on Sept 20, 2018 13:54:09 GMT -8
I **think** this is a different problem. Related perhaps. It's on our September BugWeek (which might just get pushed to the first week in October).
BP
|
|
|
|
Post by edharris on Oct 11, 2018 17:00:24 GMT -8
Bump - any update on this issue?
Thanks,
-Ed
|
|
|
|
Post by SDL on Oct 16, 2018 9:10:36 GMT -8
Ed,
Would you check to see what version of software you are running on your OurWeather? It shows up on the boot up OLED screens.
Secondly, the Pi driver needs to be rewritten. That is the conclusion we came to on GroveWeatherPi. It's on our engineering list now.
BP
|
|
|
|
Post by edharris on Oct 25, 2018 9:47:28 GMT -8
I'm running version 033.
|
|
|
|
Post by SDL on Oct 25, 2018 16:28:22 GMT -8
Ed,
We are going to have to rebuild our test system (we are using WXLink) and start a run to duplicate this. When we developed the software, we had NO issues.
Are you using the REST interface?
BP
|
|
|
|
Post by Powderjockey on Oct 27, 2018 15:06:48 GMT -8
So what is the current version of firmware for the OurWeather Kit?
|
|
|
|
Post by SDL on Oct 29, 2018 13:16:34 GMT -8
V033 on the GitHub repository.
We are rolling this out on auto update this week.
BP
|
|
nedhead
New Member
Retired EE
Posts: 15
Raspberry Pi: Yes
|
Post by nedhead on Nov 4, 2018 15:02:57 GMT -8
I, too, am having similar problems. The unit hangs anywhere from 12 to 72 hours requiring a power cycle to get it going properly. I get date/time like 85/165/2009 37:165, Outside humidity NaN, Indoor temp: -216 F, BP 53.78. When the unit hangs it stops reporting to underground weather EXCEPT for wind speed, wind gust, and wind direction - these value keep reporting. The unit is pingable, but can't provide any data past the first time I query the unit with the OurWeather page. I am running on firmware version 33. This problem has happened on 3 different Ardurino weather boards. I tried to compile 34, but it wouldn't compile. When will this be fixed ?? It is obviously a relatively common problem for other users!
|
|
|
|
Post by SDL on Nov 6, 2018 14:39:42 GMT -8
Nedhead,
You say you have used this on "3 different Arduino weather board"? What are thee boards? Model number would be appreciated.
I'm a bit at a loss on this problem since we haven't been able to reproduce it. Since the clock is screwed up too, it is even more strange. But the clock i still keeping the internal time, correct?
BP
|
|
perrito
New Member
Posts: 22
Raspberry Pi: Yes
|
Post by perrito on Dec 30, 2018 17:54:10 GMT -8
Hey did you ever manage to discover what was this? I am hitting the exact same use case but my system is in a rather inaccessible place  so hard resets require me to down the breaker for the system and our house internet link |
|
|
|
Post by SDL on Dec 31, 2018 13:58:44 GMT -8
|
|
so hard resets require me to down the breaker for the system and our house internet link