meteohub.de

Running 4.9u (Build 6942)

Hi,

Yesterday my VP2 outdoor T/H sense failed (70C & 100% RH). This morning I took the shield & sensor off the ISS for study & planned to leave it off till a replacement sensor arrives. As soon as I'd come in from the garden I checked the VP2 console to see if all the other sensors (i.e. wind, rain, solar & indoor T/H) were reporting after the disconnect/reconnect & they were all working fine.

It was only several hours later that I checked Meteohub & discovered that while rain, solar & indoor T/H were logged just fine, wind was showing as not having reported in since the time when I disconnected the outdoor T/H.

After I went out & reconnected the faulty outdoor T/H Meteohub started "seeing" the wind sensor again, so I guess I'd better leave things like that till the replacement outdoor T/H arrives.

Is there maybe some weird dependency between the two sensors that causes the wind data not (which the console was seeing fine) to be seen in the external data stream if the outdoor T/H is not present?

wind is bound to the temp sensor to get the wind chill right, which is part of the wind data.
so if windchill assigned temp sensor fails, no wind data either.

admin wrote:wind is bound to the temp sensor to get the wind chill right, which is part of the wind data.
so if windchill assigned temp sensor fails, no wind data either.

Thanks for the logical explanation.

The English idiom "Throwing out the baby with the bath water" comes to mind. I hope that translates well.

i.e. it seems a little odd to fail wind speed & direction (primary data which many might consider of high value) just because wind chill (secondary data) can't be calculated. Might it not be an alternative to set the wind chill to an flag (invalid?) value & report the issue along with other failed sensor data in the usual email?

I know that this sounds a bit strange, but coupling windchill to wind sensor and not taking it as
separate sensor is something related to the wmr928 I started with years ago and thereby
is digged very deep into the design of the logger sw. I agree that this is a design flaw.

To get it out there I would have to touch code for all weather stations and I would implement
some bugs there for sure which will keep me busy for weeks and customers complainig, why I
did ever touch that for fixing something that normally does not come to daylight. Being there a
complete overhaul of how to connect basic data sets to sensors comes into my mind, but that will
blow-up code base rather radical. May be a next generation Meteohub will have that sorted out.

admin wrote:To get it out there I would have to touch code for all weather stations and I would implement
some bugs there for sure which will keep me busy for weeks...

As a fellow software developer I know the feeling!

admin wrote:May be a next generation Meteohub will have that sorted out.

Okay. Maybe some day. Thanks for your comments.

PS. New sensor fitted now, so normal service resumed.