A water flow counter converts a spinning rotor’s pulses into flow rate and total volume, using a model-specific pulses-per-liter constant to translate signal into gallons or liters.
A small turbine sits in the pipe with magnets embedded in its blades. Water turns the rotor, and a Hall-effect sensor beside it clicks off an electrical pulse for every pass. The controller counts pulse frequency to get flow rate, counts total pulses to get volume. That’s the whole mechanism — the complexity lives in the calibration number, and getting it wrong is the mistake that quietly ruins every reading downstream.
Electronic and mechanical counters arrive at the same numbers by different roads. Here’s how each one actually works, what the published specs mean in practice, and where cheap sensors stop being trustworthy.
How Does The Rotor Turn Water Into Data?
The sensor outputs a square wave whose frequency tracks flow. A published G1/2-inch sensor spec states F = 7.5Q, where Q is liters per minute, at 450 pulses per liter. A G3/4-inch model uses F = 5.5Q and 330 pulses per liter.
Those constants are the entire conversion. Your microcontroller or counter does two calculations:
- Flow rate — measured pulses per second divided by the sensor’s pulse-frequency constant.
- Volume — total accumulated pulses divided by pulses per liter.
Drop the wrong constant in and the math still runs, still looks plausible, and is still wrong. This is why the model number matters more than the code.
Mechanical meters skip the electronics entirely. Water drives a turbine, piston, or displacement chamber, and a gear train moves the dial. No power, no calibration constant to mistype — but no data output either.
What Do The Published Specs Actually Tell You?
Range, accuracy, and pressure ratings define where a sensor works and where it starts lying. The hobby-grade units cluster tightly around the same numbers.
| Model | Flow Range & Stated Accuracy | Electrical & Connection |
|---|---|---|
| G1/2-inch sensor | 1–30 L/min; ±2%; 1.75 MPa max; 80°C max | DC 5–18 V; 10 mA max; 15 cm cable; red VCC, black ground, yellow signal |
| G3/4-inch sensor | 1–60 L/min; ±2%; 1.75 MPa max; 80°C max | DC 5–18 V; 10 mA max; 1.2 m cable; 330 pulses/L |
| G1/2-inch brass sensor | 1–25 L/min; ±3%; 1.75 MPa max | DC 5–15 V; 30 cm cable; 65 × 30 × 28 mm |
| G1/2-inch brass with temperature sensor | 1–25 L/min; ±3%; 1.75 MPa max | DC 5–15 V; temperature sensor included |
| G3/4-inch brass sensor | 2–50 L/min; ±3%; 1.75 MPa max | DC 5–15 V; 30 cm cable; 60 × 30 × 28 mm |
Read the range as a hard boundary, not a suggestion. A 1–25 L/min sensor has no accuracy claim below 1 L/min or above 25 L/min — and low-flow drip measurement is exactly where people try to push it. The 1.75 MPa figure is a maximum rating, roughly 17.5 bar, not a target to run against. And none of these listings establish suitability for potable water, hot water, chemicals, or outdoor exposure.
How Do You Install One Without Skewing The Reading?
Most bad readings are plumbing problems, not electronics problems. Shut off and depressurize the line first, then match the meter to your actual flow, pressure, temperature, fluid, and connection type before cutting anything.
Thread standards trip people up. G, BSP, BSPT, and NPT threads are not interchangeable, and adapting a BSP thread to NPT requires thread-sealing tape. King Instrument’s 7100 installation instructions call for an appropriate amount of Teflon tape on external threads and prohibit paste or stick-type sealants outright — a detail worth following even on a hobby build.
Mount the sensor in the marked flow direction. Keep straight pipe upstream and downstream, keep the meter full of water, and keep air out of the line — entrained air spins the rotor and inflates your total. Pacific Northwest National Laboratory’s water metering guidance flags air intrusion, turbulence from nearby fittings, vibration, freezing, and unsupported pipe load as measurement killers. Support the plumbing so the meter isn’t carrying pipe weight.
Wiring is tight and unforgiving. On the G1/2-inch sensor, red goes to supply, black to ground, yellow to the pulse input, and the load must not exceed 10 mA at 5 V. Once wired, load the exact model’s pulse constant, run water, check for leaks, and compare the displayed total against a measured volume — a bucket of known size is a perfectly good reference.
If you’re comparing units before buying, this roundup of tested water flow counter options covers the models worth shortlisting.
For regulated, billing, or potable-water use, none of this applies — you need a certified legal-for-trade meter for your jurisdiction. Hobby sensors aren’t that. For everything else, the official King Instrument 7100 installation instructions are a solid reference for sealing and stress-loading practice.
What Sends These Counters Wrong?
Five failures cover nearly every bad reading. The wrong pulses-per-liter constant, reversed signal wiring, exceeding the 10 mA output-load limit, overvolting the sensor, and installing in a partially filled pipe or beside a turbulent fitting.
One more deserves its own line: confusing instantaneous flow rate with cumulative volume. They’re different outputs from the same pulse train, and a display showing 12 L/min tells you nothing about how much water moved today. Check which number your controller is actually surfacing before you trust it.
Common Questions
Can a flow sensor measure total gallons, or only rate?
Both, from the same pulse signal. Rate comes from pulses per second divided by the frequency constant; total volume comes from accumulated pulses divided by pulses per liter. Most controllers display one or the other at a time, so confirm which value you’re reading before drawing conclusions about usage.
Why does my counter read high or low compared to a measured bucket?
The pulses-per-liter constant is the usual culprit. Each model has its own value — 450 pulses per liter on one common G1/2-inch sensor, 330 on a G3/4-inch version. Entering a different sensor’s number scales every reading wrong. Air in the line inflates totals too, since bubbles spin the rotor without adding real volume.
Do these sensors work with Arduino or Raspberry Pi?
Yes — they output a square wave on a low-voltage pulse line, which is exactly what a microcontroller counts. The G1/2-inch unit runs on DC 5–18 V and draws at most 10 mA, so a 5 V supply works but needs the output load kept under that limit. The listings name no required app, cloud plan, or region.
Sources
- King Instrument Company. “7100 Installation Instructions V5” Supports thread-sealing tape requirements and the prohibition on paste sealants.
- Pacific Northwest National Laboratory. “Water Metering Best Practices” Supports guidance on air intrusion, straight pipe, vibration, and installation conditions.
- Seqwater. “Flow Meters — Standard Specification” Supports installation orientation and protection requirements for flow meters.