When [Fab] needed an anemometer for his latest project, he was stymied by the limited range and relatively high prices of commercial options. Undeterred, his solution was an impressive DIY anemometer that rivals the off-the-shelf alternatives.
AnemoSens was designed from the ground up as a component for the ambitious WinDIY_2 Horizontal Axis Wind Turbine, however it’s just as suitable as part of your standard home weather station. The microcontroller unit uses RS485/Modbus connectivity, ensuring that data from the wind sensor is accessible across a variety of platforms. Serial-stream via USB and an SD cart slot are also available for recording data, the latter being particularly useful for long-term unsupervised monitoring. [Fab] also integrated an ESP32 for recording data over the air.
The MCU also features a location for the venerable BME280, which is a relatively accurate temperature, pressure and humidity sensor often deployed in DIY weather stations. This feels like a nice touch, as it means the anemometer package alone could feasibly serve as a rudimentary weather sensing station, or as a backup for more elaborate environmental monitoring.
The prototype currently uses a Hall effect sensor for measuring the wind speed, while a AS5048B magnetic rotary encoder does a decent job of measuring rotation (wind direction). Some calibration is likely necessary to improve the accuracy of this setup, but it’s a promising start.
[Fab] has already identified some shortcomings with the bearing, but has a plan for future iterations. He might want to check out this spare-parts anemometer that uses a bearing from an old hard drive.
It’s a 14bit absolute direction sensor, not much calibration needed.
The calibration part was probably about getting a wind speed (m/s) reading from the rotation speed (rpm). That’s probably a non-linear correlation, so he would need a calibrated anemometer to calibrate his DIY anemometer in the first place.
There is a comment about calibrating the anemometer by holding it out of a car and driving around the block a few times … that’s actually pretty smart :-)
Things to consider there are that the air accelerates around the car (solvable with a long stick) and getting accurate speed readings for the car (probably solvable by recording GPS data at the same time).
Glad you guys posted about calibrating as I had always wondered how that would be accomplished.
isn’t a wind turbine an anemometer itself?
Not exactly … especially with variable blades the rpm depend on the current pitch of the blades and the current wind speed and the direction of the wind and how long it takes for the massive blades to spin up … so in theory you could calculate the current wind speed from all of that (given a calibration), but it would be a relatively bad and complex anemometer that does not quickly react to changes in the wind speed.