Post by robotsquirrel on Aug 6, 2019 11:32:30 GMT -8
RobotSquirrel here, having the exact same issues again with the new board and sensor. Everything worked for a couple of days and then it failed just like the original poster here. I pulled it down from the mast and have been testing it. It stays up for about 5-15 minutes and then starts failing again.
Yep, it looks like there are two solutions here and one will require some software to be written. A new version of OurWeather later this week hopefully. I believe that we have a higher than normal failure rate on the AM2315s.
Post by robotsquirrel on Aug 7, 2019 11:07:37 GMT -8
After bringing the weather station inside to run some more tests I can't get it to fail. Yesterday it would fail just a few minutes after rebooting. When I received the new hardware I ran it inside for several days to confirm it was all working and it never failed. It was only after running outside that it failed. After running it outside it failed within a week, then consistently < 30 minutes after rebooting.
The only discernible difference I can see (other than outside temperature) is the USB power supply I am using. When I run it inside I am using an Anker Wall Charger that supplies plenty of power. When I run it outside I am using a smaller generic USB phone charger.
What's the max current I need to be able to supply to the board with temperature sensor and air quality sensor running?
You know, I believe you are onto the problem. The Air Quality sensor takes a lot of current when measuring. I'm not sure how much current everything together would take, but it might not be a current problem, it might be a power supply noise issue. Can you use the big brick outside temporarily?
Post by robotsquirrel on Aug 8, 2019 9:26:07 GMT -8
This is definitely starting to look like an issue with how the device is powered. I have tried several different configurations including which plug, different extension cords and different USB power bricks. I can get it to fail pretty quick (<15 minutes) in some of these configurations.
Anker USB power brick on a heavy duty extension cord plugged into a surge protector works without fail.
1.) Generic USB brick will cause a failure no matter what.
2.) Generic 6' extension cord causes failure even with the Anker USB.
3.) Generic 120V light bulb screw in socket adapter causes failure no matter what.
If I use a wall outlet with a solid connection; a good, heavy duty extension cord and a high quality USB power supply there are no failures. This isn't a long-term test but since I was able to fail the device quickly with some of the other configurations I think I might have isolated the cause.
One thing to note: This failure is slightly different than the original failure I was getting previously. This failure seemed to corrupt the date-time shown, which the previous failure did not. This failure seems to mirror exactly what some other users were having though.
Without the AM2315 plugged into the board it's much less temperamental with power supply choice.
BP, I was away for a bit, will be sending my board for change. Also seeing that you are doing a replacement program for the AM2315. Do you know if the replacement one you send at beginning of June one of the new batch or was it the old batch?