Driver Sample Variation, Matching, And Repeatability

Two units with the same model number are not guaranteed to be identical, and one repeated trace is not a population study.

Estimate repeatability first

Remove and reinstall one sample several times using the same fixture and state. The observed spread sets a floor below which apparent sample differences may be setup noise.

Measure a population

Record enough samples to support the intended claim. Preserve every individual curve along with summary statistics, tolerance bands, and sample identifiers.

Match for the actual role

Re, Fs, Q and impedance around the crossover
sensitivity and response over the used passband
directivity and acoustic phase when relevant
distortion, compression, rub and buzz
thermal and excursion behaviour for high-output roles

Investigate outliers

Repeat the fixture, orientation, contact, temperature, and calibration before rejecting a driver. Document whether the outlier was measurement error, reversible state, manufacturing variation, damage, or repair.

Tie samples to the system

Record which physical driver is installed on each channel, in which enclosure or location, with which crossover, gain, measurement, and approved release.