How do I account for drift in RF test equipment between calibration intervals?

How do I account for drift in RF test equipment between calibration intervals? Instrument drift is the gradual change in performance over time, and it must be quantified and included in the uncertainty budget as a Type B contribution: (1) What causes drift: component aging: the internal oscillators, references, and amplifiers gradually change their characteristics. Thermal cycling: repeated heating and cooling (power on/off cycles) stresses internal components. Humidity exposure: moisture absorption in circuit boards and components. Mechanical wear: switches, relays, and connectors in the instrument degrade with use. (2) Quantifying drift: from calibration history: compare the as-found values at each calibration to the as-left values from the previous calibration. The difference is the total drift over the calibration interval. Example: power meter reference accuracy. As-left at last calibration (Jan 2025): +0.02 dB relative to nominal. As-found at this calibration (Jan 2026): +0.05 dB relative to nominal. Drift over 12 months: 0.03 dB. If you have multiple calibration cycles: plot the drift trend over time. Calculate the maximum drift per calibration interval. From manufacturer specifications: the manufacturer often specifies drift as a ±X per year or ±X per calibration interval. This is a conservative (worst-case) estimate. Standard uncertainty: u_drift = max_drift / √3 (rectangular distribution, assuming the drift is equally likely to be anywhere within the range). (3) Including drift in the uncertainty budget: the drift contribution is added to the instrument uncertainty at the time of measurement. At the start of the calibration interval (immediately after calibration): drift contribution = 0. At the end: drift contribution = full stated drift. For a measurement made midway through the interval: drift contribution = drift × (time_since_cal / cal_interval). Conservative approach: use the full drift value regardless of when the measurement is made. This overestimates the uncertainty at the beginning of the interval but is simpler and safer. (4) Reducing drift impact: shorter calibration intervals: reduces the maximum possible drift (but increases cost). Interim checks with check standards: detect drift early (before it reaches the specification limit). Environmental control: stable temperature and humidity slow the drift mechanisms. Continuous instrument operation: leaving instruments powered on 24/7 reduces thermal cycling and associated drift.