Calibration
Calibration practice, certificates, intervals and equipment.
What is CMC in calibration?
Calibration and Measurement Capability is a laboratory's best-case uncertainty, not the accuracy of your instrument.
Dry block vs environmental chamber
Two common ways to generate a temperature for calibration, suited to different sensors, ranges and uncertainties.
How to calibrate a Pt100
Calibration by comparison against a reference thermometer, and the factors that decide the uncertainty.
As-found vs as-left
Why both sets of readings matter, and what to do when the as-found data are out of tolerance.
What should a calibration certificate contain?
A checklist for reading a certificate — and the items whose absence should make you ask questions.
Calibration curves: fitting, residuals and the uncertainty they add
Fitting the line is the easy part. Judging whether the fit is honest, and carrying its uncertainty into the budget, is where calibration curves go wrong.
How to set calibration intervals
There is no universal interval. Set it from risk, manufacturer data and — above all — the instrument's own as-found history.
How thermocouples are calibrated
Comparison against a reference in a bath or furnace, with the reference junction realised. Ignore it on a type K at 420 °C and you read 398 °C.
Data logger calibration: how temperature loggers are calibrated
A logger is calibrated as a whole system, sensor and electronics and firmware together, in a chamber, at the temperatures you actually use it at.
Liquid bath versus dry-block calibrator
With the same reference and the same repeatability, a stirred bath gives U = 0.038 °C and a dry block 0.139 °C, where the source becomes the largest term.
Immersion depth and stem conduction
A probe conducts heat along its stem. Insufficient immersion makes the sensor read closer to room temperature than to the medium.
Calibration, adjustment and verification are different things
Calibration measures, adjustment changes the instrument, verification compares against a requirement. Mixing them up loses the as-found data for good.
Tolerance versus uncertainty
Tolerance is what you require of the instrument; uncertainty is how well the calibration could measure it. A conformity statement needs both.
Choosing calibration points
Calibrate where you measure. Points should bracket the use range and include any temperature where a decision is made.
Self-heating in resistance thermometers
The measuring current warms the element above its surroundings: under a millikelvin in a stirred bath, tens in still air. Calibrate in the medium you use.
Hysteresis in temperature sensors
A sensor can read differently at the same temperature depending on where it came from. Calibrations that only approach from one direction hide it.
Selecting a reference thermometer
Pick a reference for stability first and certificate uncertainty second: one that drifts between its own calibrations is worse than a larger honest number.
The ice-point check
A properly made ice bath gives 0 °C to within a few millikelvin — a cheap, repeatable way to watch a PRT for drift between calibrations.
Calibrating relative-humidity sensors
Saturation vapour pressure climbs about 6 % per degree, so half a degree between sensor and reference costs more RH error than most whole budgets.