[laulau]

Hi,
Yesterday afternoon i was able to put the TX board outside with sensors connected.
Temperature with 10m cable works fine.

Wind vane (oregon with about 15m cable) is ok.

Wind speed (same cable as wind vane) works but speed is not accurate.
Maybe the form factor isn't well adjusted and the anemometer position isn't at the best position (1.5m above the roof). My actual anemometer is about 3.5m above the roof and give me a wind average 4 time higher and is corresponding nearly what the local met office is giving.

Rain gauge with 8m cable give me aleatory rainfall when it isn't raining.
Since yesterday night at midnight i have 6.1mm of rain that'a about 50 rain gauge clicks !
Is the internal pullup resistor sufficient or should i add an external resistor ?

Thanks to advice me.

P.S. I am glad I found this project and implement it

Link to the test page : http://laurentmey.fr/test/trends.htm

[werk_ag]

I'm impressed that it even works with that such amount of long cables connected to the TX unit.
The best advice is try to put the TX unit as close to the sensors as possible, and reduce the cable lengths.
I think you are trying to avoid put the TX unit in a box outside of the house, but maybe it's better using a single long cable to supply the power to the TX board, than having a lot of long cables. Most of the sensors are connected directly to the microcontroller pins, so many long cables can cause troubles.

Regarding the wind speed sensor I can't help more than I already did, officially it's a not supported unit, and I never tested it.

I don't know nothing about the rain gauge that you are using, usually they only have a reed switch, and nothing else, maybe yours have anything else that is causing false readings.
The rain gauge usually is the unit less sensible to cable length.

Why not use the tested Fine Offset instruments? They are cheap, and easy to find, at least in Europe.

[laulau]

I think you are trying to avoid put the TX unit in a box outside of the house, ...

Not really.
I plan to built two Tx boards for the final implantation so the cables will be shorter
Actually it's just a test and the cables of the rain gauge and the wind sensor are the original ones.
The rain gauge has only a reed switch but as it's a long time i didn't use it i have to check if the connexions are clean.
In the future I'm thinking to buy a Davis wind sensor and perhaps an other rain gauge.

P.S.
I added a 10k resistor around 14H, rainfall was stable until 23:45. Then again the false rain readingss. This may come from moisture, i'll check it tomorrow.

[laulau]

The false rain readings seem to have stopped since i clean the rain gauge and especially the little circuit (with the reed switch) which was oxidized.
I've soldered the wires directly on the reed switch.
I'm now waiting for real rain to see if it's working well.

[werk_ag]

Good to know. Thanks for the update

[laulau]

It rained this night !
   
The rain rate dropped to zero at about 3:00. i wonder why as it was continuously raining until 3:45.
How is rain rate calculated ?

[werk_ag]

Looking at the graph I don't see abnormal problem.

Rain rate is calculated based on the time between successive tips of the rain collector.
If there have been no rain tips since the last rain rate data transmission, then the rain rate based on the time since that last tip is indicated. This results in slowly decaying rate values, instead of showing a rain rate which drops suddenly to zero. This results in a more realistic representation of the actual rain event.
If this time exceeds a defined amount of time, than the rain rate value is reset to zero.
As you can see on your graph, despite you say that was continuously raining, between 2:50 and 3:00 there isn't any rain tip, this causes the reset of the rain rate to zero.
This will happen many times during low rain fall rate periods. If you want improve this, you need to improve the quality of your rain gauge to a unit of 0.1mm per tip, and even so a zero rain fail rate can be computed many times under very low rain fall rates.

There isn't a standard for the rain fall rate calculation, the WeatherDuino Pro2 try to follow the Davis approach.

I was looking to the graphs in your website, which I think are still from your other weather station, and I saw that the rainfall rate is only reset to zero, two hours (3:49 - 5:59) after the last rain tip registered.
As said there isn't a standard for rainfall rate presentation, but two hours seems very excessive.

...If this time exceeds roughly 15 minutes, than the rain rate value is reset to zero. This period of time was chosen because 15 minutes is defined by the U.S. National Weather Service as intervening time upon which one rain "event" is considered separate from another rain "event".

[laulau]

Thanks for the explanation.
I was asking because the rain rate graph was very different from the one from the WMR928 station data.

[werk_ag]

It was what I thought.
No problem, questions are welcome.
If you Google a little, there are a lot of interesting discussions about the rainfall rate calculation.
Normally, the WeatherDuino Pro2 does almost all calculations at the RX level, except the Rainfall rate that is calculated by the TX unit. Davis does this exactly the same way, because we need a way to accurately measure the time between tips.

Rainfall rate reported is an instantaneous calculation, extrapolated to an hour.