A pulse is not a speed. Treat the cup, the switch and the mount as one measurement chain whose calibration you have to establish, and keep the instrument’s role modest: hobby logging, not a storm-warning device.
Wind-sensing hardware has two parts that are often described as one: the mechanical rotor that turns airflow into rotation, and the switch that turns rotation into electrical events. Everything downstream depends on both, so it pays to state what each contributes before choosing a sensor. The software side of the same chain is covered in the companion anemometer software data-quality guide.
Events are not automatically wind speed
A reed switch or a Hall-effect sensor produces transitions. A transition tells you something moved; it does not tell you the airspeed. Turning one into the other depends on a documented relationship between rotation and speed, and that relationship is a calibration question, not a wiring question. Cup geometry on its own does not verify wind-speed calibration.
Interrupt handling assumptions
The Arduino attachInterrupt reference lists pins 2 and 3 as the interrupt-capable pins on the UNO Rev3, notes that pin mappings vary across boards, and recommends digitalPinToInterrupt(pin) together with keeping interrupt service routines short. Those are design constraints rather than afterthoughts: a long routine inside an interrupt can stall the counting you depend on.
What to document for each sensor path
| Question | Contact switch (e.g. reed) | Solid-state sensor (e.g. Hall) |
|---|---|---|
| Event source | Mechanical contact closure | Magnetic field sensing |
| Qualification need | Contact bounce can multiply one transition | Confirm the specific part output behaviour |
| Electrical interface | Check the pull-up and current it needs | Confirm supply and output type from its own datasheet |
| Mechanical concerns | Moving contacts and magnets wear | Check mounting and magnet retention |
| Calibration input | The rotation-to-event ratio | The rotation-to-event ratio |
The debounce example’s point, that elapsed-time checking qualifies a changed state and that bounce otherwise multiplies transitions, applies to any contact-based approach, and it is a hardware-facing assumption as much as a software one.
An ordered evaluation process
- State the measurement target. Fix the speed range and averaging you actually care about.
- Identify the real part. Read the sensor’s own datasheet rather than inferring from a similar-sounding number.
- Establish the rotation-to-event ratio. This is the bridge from events to speed and must be documented.
- Plan calibration against a reference. Decide what known reference the reading will be compared with, and when.
- Check electrical compatibility. Voltage, pull-up and output type come from the exact part document.
- Plan for exposure. Friction, bearing drag, magnet retention, moisture and vibration are mechanical questions that shape the reading.
- Set the qualification strategy. Decide at the switch how many transitions count as one.
- Record the limits. Write down the calibration range and conditions the instrument is valid for.
Reader questions
Why can I not just count pulses and print a speed?
Because the conversion depends on a documented rotation-to-event ratio and a chosen averaging window. Without both, the number is a count, not a speed.
Does contact bounce really change the result?
Yes. The debounce example documents that one mechanical transition can become several state changes, and each extra count inflates the derived value.
Do all boards accept interrupts on the same pins?
No. The reference lists pins 2 and 3 for the UNO Rev3 and notes that mappings vary across boards, so check your own board.
Is this a weather-grade instrument?
No. Treat it as hobby logging and evaluate mechanical security and electrical compatibility independently; it is not a safety-critical wind instrument.
Documentation
- Arduino attachInterrupt() reference (official reference-en source)
- Arduino Debounce example (official arduino-examples source)
- ESP8266 Arduino Core official Reference
Scope note: these links cover the interrupt reference, the debounce example and the platform core reference only. No build or bench measurement was performed for this guide, so actual part numbers, input-voltage compatibility, magnet retention and environmental durability still need to be identified.
