Anyone figured out best way to detect rain?

I remember last time i asked, there was some weather station ways to detect rain and wind.

But those items weren't zwave and kinda bulky.

Anything out now that can sense rain (a must) and wind (optional)? Zwave preferred.

Zigbee. Will this work? put it onto a small plate outside. When it rains, collects water, triggers. When sun is out, water dries. Ready to fire again.

I think detecting rain is easy, the issue is detecting when the rain stops and keeping the sensor clean.

eBay has a number of rain detector boards that consist of a simple circuit board with traces interlaced. Connect one of these to any leak detector and you have a rain detector.

The issues will be:

  • Keeping it clean
  • Water retention after the rain stops
  • Snow may or may not trigger it.
2 Likes

Tempest

4 Likes

ECOWITT WH40 is a mechanical range gauge which needs a gateway. Total cost about $85.

I expect it depends on the application, i.e. what you are wanting to do once you detect rainfall. I'm no expert, but I believe some of the differences in what people want and see in rainfall detection are due to the fact many weather stations require a certain level of rainfall to have occurred before they start reporting, particularly the older-style sensors. For some weather stations it is more about recording the amount of rainfall like a traditional rain gauge would, rather than detecting and reporting the instant that it starts raining. Some people need/want to detect rainfall much earlier and after a much smaller amount of rain, which is where some of the other solutions come into play, such as detection pads of some kind (I really haven't researched this :slight_smile: ).

I don't have any solutions as such, just acknowledging that there are differences between traditional weather stations and other solutions where detection of smaller and smaller rainfall events is important, and that identifying the intended outcome is important in choosing the right sensor, not just the communication protocol or the fact that a device detects rainfall.

I can understand the desire for Zwave. While I have not used it myself for any of my setup, I believe it has a stronger signal that can penetrate through more varied materials and does not suffer from the interference issues that Zigbee and Wi-Fi devices can. I also believe, and may be wrong about this, that some of the frequencies used in many weather stations would also have some similar properties.

I wont say this is the best way, but it is the best way for me. I wired a self emptying rain meter to a zwave contact sensor. The emptying of the rain cup causes the contact sensor to open/close.

I use this automation to control a virtual switch, on = raining, off = not raining

Rain Meter:
https://www.amazon.com/gp/product/B00QDMBXUA/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

Zwave contact sensor with external leads:
https://www.amazon.com/gp/product/B08LDJJWSY/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&th=1

Either way that’s my 100% local way to detect rain thats been 100% reliable for years.

1 Like

I think that the Aeotec / Samsung sensor would work. I suggest that you try it positioned at a 45 degree angle with the bottom electrodes facing up on the lower edge. They are close together and should be triggered by just a few raindrops. The water would run off. No need for a plate. I'm not sure how well the device will last with full outdoor exposure. You could try painting over the seam between the two pieces with rubber cement and pealing it off to change the battery.

1 Like

While a regular "leak sensor" would detect rain quickly, the reset time, after the rain ended, would be unpredictable. I also think most leak sensors have built in audible alarms, which could be annoying. Also many of them are not designed to work when submersed in water, which could happen in heavy rainfall.

Underground irrigation systems all make use of a rain sensor. The one I have can detect moderate rainfall within a few minutes, and will reset itself within an hour of the rain stopping. The reset time depends on wind speed, temperature, and sunlight. These sensors are a simple mechanical switch (no electronics).

My suggestion is to connect the irrigation rain sensor to a simple Z-Wave door/window sensor that has external connection points. Of course the contact sensor would have to be placed in a protected area, and connected to the rain sensor with a few feet of thin speaker wire.

If you're ok with Zigbee, maybe take a look at this forum.

Edit: It's been a while since I visited this forum, and it seems like some users are experiencing connection issues. Maybe it's not quite ready for prime time yet.

If there is heavy fog or mist that makes everything wet to the touch... But won't fill a tipping cup gauge ... Does that count as rain? What surfaces are covered in water, but it comes from condensation and not the sky? Should that set off the rain sensor?

Regardless, put down another vote for Ecowitt stuff. It's affordable, reliable, and fun.

Just buy ecowitt already... i use it to send mu automower back to base on rain.

Until I finally went with Ecowitt WS90 Multisensor all may attempt to use leak sensors, etc. were semi-successfull. I.e. DIY toy worked for a short time and than failed.

Just put a cam under a glass bowl pointing up.
Look at the image.
Problem solved!

3 Likes

Sure, very nice solution with involving the entire Internet Infrastructure plus AI.

3 Likes

looks like ecowitt is the way to go. I want something reliable and sensitive in terms of detecting rain.

I see 2 good models.

WH40

WS90

WH40 is more accurate when it comes to rain. WS90 description says "The daily rain deviation of WS90 is very small in the long run, but under certain conditions, the deviation could be larger: As the raindrop size and wind speed can have different impacts on the sensor output which leads to this variance. The WS90 product suffers from this imperfectness. If you are very demanding on rain data accuracy, we recommend you purchase WH40 and use it with WS90. If there is no precise requirement on data of each rain, then WS90 is just fine."

My use case is simple. Detect rain, close alfressco blinds.
I could later tinker with it, and possibly make the blinds reopen when 10 minutes of no rain has passed.

Looks like i need to get ecowitt gateway as well.

the own app shows alot of useful information.

do these information feed into hubitat? would love to do show the data on my custom dashboards and do some fancy logic on it.

how good is it in detecting rain? does it need alot to trigger?

1 Like

this really depends on how you position it. Rain cups are simple devices and used by meteorological services for the same purpose, They set the standard at detecting and measuring rain fall.

The Amazon link shows the size of the cup, that’s exactly how much it needs to trigger.

Yes, you can use the driver I now look after:

Yes and no.... You need something to act as a gateway, which will often be the kind of device you showed, but some of the displays EcoWitt sell (not all, I believe) can perform the same function, but check the displays features before purchasing.

The gateway is the device that the sensors send their readings to and it's the gateway that handles sending that data on to Hubitat and other destinations that you choose.

mine is not immediate but with a light drizzle does within 5-10 mins

In my case the WS90 location is too far from perfect - balcony rail on a 9th floor in a BIG apartment building.
Newer the less and to my BIG surprise, Wind and Rain detection is near perfect.
Light, Temperature and Humidity detection is ideal. All parameters from WS90 are used in different RM Rules. So far so good - all the related rules are doing their job almost ideally.
For the Wind and Rain detection it is even better than my expectation was.

Tempest Weatherflow is also pretty awesome.