I like measuring things, and I like the weather. So why shouldn't I own a weather station? I've thought about it for several years, but never committed to it. Why not? Cost! But since I found a $35 solution, this year I did it!
 |
| My Weather Station Configuration |
Why did it take me so long?
When I looked online, I saw a bunch of $120 to $1000+ kits that did
everything: wind, rain, solar radiation, barometric pressure,
temperature, humidity, lightning, etc. They looked big and complicated, and was more than I wanted to spend on what amounts to be a toy.
The reasons I didn't buy one of these weather stations:
- Installing a weather station seemed like a pain in the butt.
- My property isn't really set up for quality weather data collection (limited open space, with trees, buildings, fences, etc).
- I didn't want to spend a lot of cash on finicky equipment.
- Every other weather-related thing I've ever purchased has been pretty limited.
But then I learned that this isn't the only way. The trick is leveraging the ecosystems.
Ecosystems versus Kits
Although the kits are definitely the right way to go if you want to measure everything imaginable all at once, there is another way: Don't buy a kit, and instead by the individual units: There are sensors, and there are receivers, and they work within an ecosystem. The kits are simply a bundles of sensors with a receiver.
So I instead went down the path of buying the sensors I wanted, and leaving the rest out.
My biggest decision was to stick with a "displayless" weather station, and instead just have my data displayed on a computer, phone, or web page. Yes, not as flashy as one of those gimmicky "weather displays" you see in the ads, but infinitely more flexible.
Enter the Ecowitt
Ecowitt is a brand of weather sensors. They sell receivers and transmitters, so my first goal was to buy a receiver.
The GW1100 is a simple Weather WiFi Gateway: It's a displayless WiFi-connected receiver with built-in barometer and temperature and humidity sensors. It also wirelessly (not WiFi) receives and processes signals from remote sensors. The GW1100 gateway costs about $20. It is an indoor device, so the humidity and temperature it reports is "indoor" (barometric pressure should be the same inside and out). According to what I've read, the GW1100 can handle up to 64 sensor units.
The GW1100 is powered via USB, so it does need to be plugged in 24x365. I installed my GW1100 in my attic. We'll see how it survives the crazy hot and cold temperatures up there, but so far, so good. I power my GW1100 with an old Apple iPhone USB charger, and measuring its power draw with my Emporia Vue shows that it consumes about 0.6 watts of power (almost certainly less than $1 in power per year).
For about $15, I added the WH32 (also known as the WN32) outdoor sensor, which measures both temperature and humidity. This outdoor sensor isn't technically water proof (which is typical for the industry) so I keep this battery operated sensor in a protected area about 20 feet away from my house. If necessary, it could be installed within a weather proof solar shield.
I installed Lithium AA cells for long life even in the coldest, sub-freezing weather.
Configuration
It is very easy to hook up the GW1100 to personal weather services, either something like Weather Underground or your own service running on a Raspberry Pi (or whatever). It is so easy that I won't go into it here, but suffice to say that the GW1100 can send updated sensor data at a high frequency to Weather Underground and other services. Sending fresh data every 60 seconds is completely reasonable. I had the GW1100 reliably sending data in under 5 minutes. I currently send my data to Weather Underground and to Ecowitt's own service. Having data sent to my home server (a Raspberry Pi) is equally as easy.
The one weird thing I learned
I can have many temperature sensors connected to the GW1100, but only one of the "outdoor" (WN32 / WH32) variety. I think the principle is that "a PWS can only collect and report on one outdoor temperature/humidity device at a time".
The indoor sensors (WN31) are fundamentally the same, but are not eligible to send their data as outdoor weather to services. Evidently there is a technical way to convert a WN31 to a WN32, so look up that if you're interested.
Expandability
I haven't expanded the system much yet, but it is very easy to add many more sensors. With sensors available from about $10 to $200, needless to say there is something for almost everyone. I did end up buying a WN31 indoor sensor for about $10, which I use to monitor the humidity in my unfinished basement. Weather Underground ignores this sensor, but EcoWitt and my own Raspberry Pi can collect and display it.
Conclusion
That's it! $35, can't go wrong. I'd do it again in without reservations.