Hi,
Brand new here and hoping my total lack of understanding of this new topic is looked upon favourably.
I am wanting to “build” a wind speed and direction data logger. It must have the following characteristics:
wind sensor
wind direction indicator
log data, possibly every 5-10 minutes. Capacity to hold at least a few months data.
stand alone power supply - no AC available. Possibly a powerbank or DC cells capable of at least a few months without charging/changing over batteries
connect to PC to upload data
From my brief reading and YouTube foray into this new field, my impression is that a Raspberry Pi would be a bit overkill for this application. I then looked at Arduino models which got my head spinning with all the different models - Nano, Uno, Mega (apparently better due to the increased memory capacity)…
The following is an anemometer I have found but:
is it appropriate?
how can wind direction be incorporated?
will it do the job of my brief above?
I’m just after some guidance very much in layman’s terms initially.
To get the windspeed I would probably look to make something along the lines of this sensor using hall effect sensors as they don’t provide any additional mechanical resistance: https://learn.sparkfun.com/tutorials/weather-meter-hookup-guide#hardware-overview
Designing the parts so that they are rainproof might be a bit difficult but certainly not impossible!
The Anemometer above doesn’t have any power draw listed on the page or datasheet so you might have to get fancy with the way everything is connected (A FET of some sort to cut the power to the sensors).
Some parts that might be worth taking a look into
For lower power projects that have very low bandwidth outputs I would take a look at using LoRaWAN to send the information. The LoPy4 has an ESP32 at the core and can use very small amounts of current while sleeping (almost necessary to be able to deploy something in isolation for months/years)
You might even be able to add a small solar cell to keep everything topped up
One aspect of your project that we haven’t touched on yet is power. There’s a few great guides on this site about using solar and batteries for IoT projects:
LiPos have good self-discharge as opposed to Lead-Acid, but Lead-Acid is going to be cheaper per Wh generally, and it’s large size and weight is not much of a concern for a fixed installation. Lead-Acid is also a lot simpler to charge without a specialised board like the ones layed out in the guides above.
You’re right in picking a microcontroller as opposed to a SBC like a Pi, the micro will use a fraction of the power and be well and truly up to your tasks.
As far as wind direction, if you’re happy to fabricate the “flag” that is pushed by the wind, you’d just need an absolute rotary encoder to communicate the direction to the micro, or maybe an array of Hall Effect sensors and a magnet on your vane.
Core also sells a weather station with direction and speed sensing, you can just ignore the rain guage
I was just taking a look through the recent products and came across this one: https://core-electronics.com.au/rs485-wind-direction-transmitter.html
It looks like Liam’s hunch was correct with using some sort of hall sensors to read the direction.
Making something in Fusion or another modelling program could be a cool learning experience!!
