Power Meter Crankset

Overview

A power meter crankset measures the mechanical power a cyclist delivers, in watts, in real time, by instrumenting the crank itself as a torque transducer. Power is torque multiplied by angular velocity; the crank already carries all of the rider's torque and rotates at a measurable cadence, which is why the crank and its spider became the dominant measurement site. The principle is the same as an industrial torque cell: bond Foil Strain Gauge to metal where load produces predictable elastic strain, read the resistance change, and convert. What distinguishes a cycling power meter is doing this to ±1–2 % on a component that is hammered by weather, pressure-washed, and expected to run a year on a coin cell.

Measuring torque with strain

The Strain Gauge Package is the measurement core. Foil gauges, metal grids whose resistance changes about 0.2 % at full pedaling load, are bonded with heat-cured Gauge Adhesive to the Instrumented Spider or to the crank arms, in matched pairs oriented at ±45° to the axis. That geometry is the trick: torsional shear strain adds across the pair while bending strain and thermal expansion cancel. Wired as full Wheatstone bridges through the Bridge Wiring, the gauges produce a differential signal of a few millivolts at full scale that is inherently immune to supply drift.

The structure being measured is the Crank Arms & Spider itself, the spider or arm is simultaneously the drivetrain lever and the sensing spring, so its stiffness is part of the calibration and every unit is factory-characterized with known masses to establish its individual slope in N·m per volt. Spider-based designs gauge the Instrumented Spider because all chain torque passes through it regardless of which leg produced it; arm-based designs gauge the Non-Drive Arm (single-sided, doubling left power) or both arms for true left/right balance. A Temperature Sensor feeds firmware compensation for the residual thermal drift the bridge geometry cannot cancel; the pre-ride zero-offset the rider performs handles whatever remains.

From millivolts to watts

The Electronics Pod amplifies the bridge output in a low-drift Instrumentation Amplifier, digitizes it with a 24-bit 24-bit ADC at 50–200 Hz, and pairs it with cadence from the Cadence Accelerometer, a MEMS accelerometer that watches the gravity vector rotate once per revolution, eliminating the frame-mounted magnet of earlier designs. The Microcontroller integrates torque over each revolution, multiplies by angular velocity, and hands per-revolution power to the ANT+/BLE Radio, which broadcasts the ANT+ power profile and the BLE Cycling Power Service simultaneously at 4 Hz to head units and trainers. The whole pod lives in an IPX7 Pod Housing molded onto the spider or arm.

Power for all of this comes from the Battery System: classically a CR2032 Cell CR2032 behind an O-ring-sealed Battery Door, lasting 100–400 hours because the radio duty cycle, not the measurement, dominates the budget. A micropower Voltage Regulator holds the bridge excitation steady as the cell sags from 3.0 V toward 2.0 V; rechargeable variants swap the door for magnetic Charge Contacts so the seal is never broken.

Mechanical platform

Below the instrumentation, the product is a conventional high-end crankset. The Chainring Set bolts 50–53T and 34–39T rings to the instrumented spider with Chainring Bolts at about 12 N·m; because measurement happens upstream of the rings, spider-based meters tolerate ring swaps without recalibration. The Spindle & BB Interface passes a hollow 24–30 mm Spindle through the frame's Bottom Bracket, preloaded by the Preload Collar and clamped by the Crank Fixing Bolt at 35–50 N·m. One subtlety hides here: bottom bracket drag is downstream of a crank meter but upstream of a rear-hub meter, which is why crank-based units consistently read 1–2 % higher than hub-based ones on the same ride, both are correct about what they measure.

Accuracy in practice

The error budget is dominated by three terms. Slope stability depends on the Gauge Adhesive bond, a creeping epoxy joint is the canonical long-term drift mechanism. Thermal transients matter on fast descents, where the metal cools faster than the compensation can track; riders are told to re-zero after large temperature swings. And single-sided meters carry a systematic error equal to the rider's actual leg imbalance, commonly 2–6 %, which no calibration can remove. Within those limits, a maintained crank meter holds ±1–2 %, tight enough that training zones, FTP tests, and race pacing are all built on its numbers.

The category's significance is larger than the hardware: the power meter converted cycling training from speed- and heart-rate-based estimation into direct measurement of work, and the crank-based strain gauge design, first commercialized by SRM in the late 1980s, is still the reference architecture every newer approach (pedal spindles, hub torsion, optical deflection) is validated against.