Post by Johan on Jul 27, 2017 1:51:13 GMT -8
I have successfully updated the firmware on my Ourweather unit - and was very thankful for the OTA functionality, because it lives in a weatherproof box on a pole on the roof. This morning I could immediately see that ever since the update the wind speed readings were much less noisy, as hoped/expected/intended.
So that is good.
However, the wind direction may be somewhat broken now - I suspect there is a (not)wrapping error, because almost (but not quite) all the readings are in the 'lower' half of the graph (0° to 180°), with a definite bunching up of values at the 0° direction (see station IPRETORI29 on Weather Underground). Which is of course not completely impossible, given that I have, until this moment of writing, only 18 hours of data. So I looked at ICRANBRO8, where I can see that the upgrade to v022 was made on the 22nd of July at about 15h20. In the subsequent five days, all wind direction readings were between 0° and 157°, with distinct peaks at both values, which transcends any reasonable assessment of real-world probabilities, especially when seen in conjunction with the results of my limited testing of the same code on the other side of the world. At least my readings occasionally go up to 202°.
So that is good.
However, the wind direction may be somewhat broken now - I suspect there is a (not)wrapping error, because almost (but not quite) all the readings are in the 'lower' half of the graph (0° to 180°), with a definite bunching up of values at the 0° direction (see station IPRETORI29 on Weather Underground). Which is of course not completely impossible, given that I have, until this moment of writing, only 18 hours of data. So I looked at ICRANBRO8, where I can see that the upgrade to v022 was made on the 22nd of July at about 15h20. In the subsequent five days, all wind direction readings were between 0° and 157°, with distinct peaks at both values, which transcends any reasonable assessment of real-world probabilities, especially when seen in conjunction with the results of my limited testing of the same code on the other side of the world. At least my readings occasionally go up to 202°.
