|
|
Post by SDL on May 19, 2017 15:35:23 GMT -8
|
|
|
|
Post by stevearmitage on Oct 24, 2017 0:23:38 GMT -8
|
|
|
|
Post by SDL on Oct 24, 2017 8:51:29 GMT -8
Looking good guys!
BP
|
|
|
|
Post by johanv on Oct 25, 2017 5:21:07 GMT -8
With other OurWeather units on WeatherUnderground I could now at last confirm that an issue that I have been observing on my station is not unique. Occasionally the temperature and dew point will show an upwards offset - you can see this on my station: IPRETORI29 - see for example October 25, 2017  but also on others: IPERTH93 - see e.g. October 18, 2017 KMTCOLUM27 - see e.g. September 20, 2017 What I find interesting is that it is not a complete saturation - the values still follow the expected trajectories, just much higher than they should be. I do seem to find that it happens more now that summer is back again here in the southern hemisphere - but the problem does not start at the same temperature every time. I have a pretty decent sunshade, well-ventilated, with forced circulation, courtesy of a 12V computer fan (slowly) running at 5V at the bottom gently wafting air upwards to where the sensor lives, and out through slits just below the top skirt.  The picture was taken just before sunset - you can seen the fan as the black square at the bottom of the assembly on the right. It really seems as if the problem is in the electrical or software domain - if the temperature at the sensor really changed like that there is no way that it would jump up and down as suddenly and repeatably as it does. |
|
|
|
Post by SDL on Oct 25, 2017 8:31:53 GMT -8
Hmm. Sounds like noise to me on the I2C bus. However, what is the exact value that it jumps up to? I'm want the difference between the lower and upper value in the jumps.
Seems to me I remember something like this a year ago.
There are a couple of things we could suggest, including a software fix (if it jumps up, then filter it out) that is kind of a kludge.
Best regards,
BP
|
|
|
|
Post by johanv on Oct 26, 2017 1:34:13 GMT -8
I had a look at all three our stations, and, where the jumps occur, it averages out to exactly 16°C in magnitude, on all three stations. The readings are taken several minutes apart, in which time the real temperature can change slightly, so the actual jump may appear slightly larger or smaller than 16, depending on whether the temperature was rising or falling. Interestingly, in my case it only seems to happen when the actual temperature is in the thirties (Celsius, that is), while the station in Australia and the one in the US at least sometimes exhibit the same behaviour at much lower actual temperatures.
So, the jump seems to be by 16, and the error always seems to be upwards. I agree that it appears to be noise, but I do not know why it should only affect the same bit every time, in the same way.
|
|
|
|
Post by SDL on Oct 27, 2017 15:06:21 GMT -8
Johanv,
I will try to find my work when I ran across this a couple of years ago. We have been using these AM2315 sensors for years. I think it was on an ESP8266 so I will dig around in the archives tomorrow.
BP
|
|
|
|
Post by stevearmitage on Oct 30, 2017 21:05:58 GMT -8
|
|
|
|
Post by SDL on Nov 1, 2017 18:13:24 GMT -8
|
|
|
|
Post by stevearmitage on Nov 1, 2017 21:57:06 GMT -8
I was premature. www.wunderground.com/personal-weather-station/dashboard?ID=IPERTH93I've created a second weather station posting the same information The information is posted from PHP every 5 mins after storing in my DB www.wunderground.com/personal-weather-station/dashboard?ID=IPERTH115I keep trying to set the information and assume it's correct It seems to log one and then did not send any more data. I'm not sure where go to now to identify the issue A REST call to return the WU status - ID, key, last WU post date/time since last WU post and last WU post success/fail would be helpful to diagnose. I brought this to learn about programming these devices, but i'm not there, yet... Steve |
|
|
|
Post by SDL on Nov 3, 2017 8:59:57 GMT -8
Steve,
We will add your suggestion to the next release.
BP
|
|
