Measure wind speed with a 3D-printed anemometer

Optional

This lesson has free STL files. Print the parts yourself with one of these 3D printers.

Measure wind speed with a 3D-printed anemometer – EEMENTOR STEM™ lesson

Weather stations measure wind speed with a cup anemometer: three or four cups on arms that the wind spins round. The faster the wind, the faster it turns. In this lesson you print one, count its turns with the photo-interrupter and turn the count into km/h. You also learn about interrupts, a way for the UNO to notice very short events even when it is busy doing something else.

What you'll learn

How to count fast pulses with an interrupt
How to turn pulses into turns per second, and turns into a speed
Why measuring instruments need calibrating
How a slotted disc, called an encoder disc, measures rotation

What you need

UNO R3 board and USB cable (UNO R3 Starter Kit)
Photo-interrupter module (45-in-1 Sensor Kit)
3–6 V DC motor, used only as a bearing (UNO R3 Starter Kit)
3 female-to-male jumper wires
A fan or a hair dryer on its cold setting to test with
The printed parts from the 3D printing box below (disc, cups and mast), or the cardboard version described there
Breadboard and jumper wires (UNO R3 Starter Kit)

3D printing · free STL files

Print the anemometer

There are three parts. The slotted disc presses onto the shaft of the DC motor, which here is only a smooth bearing and is not connected. The cups sit on top of the disc's tube. The mast holds the motor and has a post for the photo-interrupter, so the disc passes through the sensor's slot.

⬇ Encoder disc STL · 56 × 56 × 22 mm ⬇ Cup rotor STL · 105 × 115 × 26 mm ⬇ Mast STL · 90 × 90 × 72 mm

Print settings and how to put it together

Print settings: PLA, 0.2 mm layers, 3 walls, 15 % infill. All parts print as they lie in the file, without supports. Print the cups slowly (40 mm/s) so the thin walls come out clean.

Push the motor into the top of the mast, shaft up, with its two terminals over the side window. It should be a firm fit.
Press the disc onto the shaft until it is about 2 mm above the top of the mast. Spin it with your finger: it should turn freely and wobble as little as possible.
Push the cups onto the spigot at the top of the disc's tube, with a drop of glue.
Hold the photo-interrupter module against the flat side of the sensor post, with the slot facing the disc and one arm above and one below it. Turn the rotor slowly and slide the module in until the disc runs through the middle of the slot without touching.
Fix the module there with hot glue. Connect it with three female-to-male jumper wires.

No 3D printer? Make a rotor from three halves of plastic cups glued to straws on a card disc. Cut four windows near the edge of the disc and push it onto the motor shaft.

Wiring

From	To
Photo-interrupter S	UNO pin 2
Photo-interrupter +	5V
Photo-interrupter −	GND
DC motor	Not connected

Sensor modules from different makers can label and order their pins differently. Always follow the labels printed on your module.

The sensor must be on pin 2 or 3: those are the two UNO pins that can trigger an interrupt.

The sketch

const int SENSOR_PIN = 2;
const int SLOTS = 4;                 // windows in the rotor's disc
const float CUP_RADIUS = 0.05;       // metres from the shaft to the middle of a cup
const float FACTOR = 2.5;            // wind speed / cup speed; set it when you calibrate

volatile unsigned long pulses = 0;
volatile unsigned long lastPulseUs = 0;

void countPulse() {
  unsigned long now = micros();
  if (now - lastPulseUs > 2000) {    // ignore bounces shorter than 2 ms
    pulses++;
    lastPulseUs = now;
  }
}

void setup() {
  Serial.begin(9600);
  pinMode(SENSOR_PIN, INPUT);
  attachInterrupt(digitalPinToInterrupt(SENSOR_PIN), countPulse, RISING);
  Serial.println("turns/s\tkm/h");
}

void loop() {
  delay(1000);                       // count for one second

  noInterrupts();                    // copy and reset the counter safely
  unsigned long count = pulses;
  pulses = 0;
  interrupts();

  float turnsPerSecond = (float)count / SLOTS;
  float cupSpeed = turnsPerSecond * 2 * PI * CUP_RADIUS;   // metres per second
  float windKmh = cupSpeed * FACTOR * 3.6;

  Serial.print(turnsPerSecond, 2);
  Serial.print("\t");
  Serial.println(windKmh, 1);
}

Test and calibrate

Upload the sketch, open the Serial Monitor at 9600 baud and turn the rotor slowly by hand. Every quarter turn adds one pulse.
Blow on the cups or hold the anemometer in front of a fan. Open Tools → Serial Plotter to see the speed as a graph.
To calibrate, you need a known wind. On a calm day, an adult can hold the anemometer out of a car window (passenger side, never the driver) at a steady 10, 20 and 30 km/h on a quiet road. Compare the readings with the car's speed and change FACTOR until they match.
Real anemometers are calibrated in a wind tunnel. Without one, your readings are a good estimate, not an exact value.

How it works

An interrupt makes the UNO stop whatever it is doing the moment pin 2 goes from LOW to HIGH, run countPulse() and then carry on. That way no pulse is missed, even during delay(). Variables that an interrupt changes are marked volatile, and the main loop switches interrupts off for a moment while it copies and resets the counter, so the number can't change halfway through.

Each turn of the rotor gives 4 pulses, because the disc has 4 windows. The middle of each cup travels round a circle of 2 × π × 0.05 m = 0.31 m per turn. The wind moves faster than the cups, because the cups also catch some wind on their back side; for cup anemometers the wind is usually 2 to 3 times faster than the cups. That is the FACTOR.

Try this next

Show the speed on the 1602 LCD, with the highest gust since you switched on.
Convert km/h to the Beaufort scale (0 to 12) and print names like "fresh breeze".
Move it to the ESP32 and send the wind speed to ThingSpeak, like in the cloud weather logger.

Troubleshooting

The count doesn't change: watch the module's LED as you turn the rotor slowly. If it never changes, the disc doesn't pass through the middle of the slot; move the module.
Far too many pulses: the disc wobbles and touches the slot. Push the rotor straight onto the shaft, or make the 2000 in countPulse() bigger.
It only spins in strong wind: the motor's brushes add a little friction. Spin it by hand for a minute to run it in, and keep the cups clean of strings.
A printed part is too tight or too loose: printers differ by a few tenths of a millimetre. Sand or file a tight hole, and fix a loose part with a drop of hot glue or a strip of tape.