Calibration and traceability are about being able to show where a number came from. A sticker on the case records the date somebody last compared the instrument against a reference. It does not promise the instrument is accurate now, and on its own it does not even tell you whether it was accurate when it arrived at the lab.
What calibration actually is
Calibration is a comparison. A known reference is applied to the instrument, the instrument’s response is recorded, and the difference is documented. That is the whole operation. Adjustment, meaning physically or digitally correcting the instrument so it reads closer to the reference, is a separate step that may or may not be performed. A calibration certificate that reports only post-adjustment results tells you the instrument was made to read correctly on that day. It tells you nothing about the measurements you made in the preceding year with an instrument that may have been drifting the entire time. This is why laboratories distinguish as-found data, recorded before any adjustment, from as-left data recorded afterward, and why the as-found figures are the ones with diagnostic value.
Traceability means an unbroken chain
Traceability is the property that the reference used in that comparison was itself compared against a better reference, and so on up a documented chain to a national or international standard, with the uncertainty stated at each step. It is the mechanism that lets two laboratories on different continents agree about what a volt is. A certificate claiming traceability should identify the standards used and state the uncertainty of the calibration itself, because a comparison is only as good as the reference behind it. Calibrating a good instrument against a reference barely better than it is produces a certificate with an uncertainty so large that it cannot confirm the specification. When a result must be defended, the uncertainty line on the certificate matters as much as the deviation figures.
Why the as-found data is the useful part
When an instrument comes back showing it was reading within specification before adjustment, you have evidence that the measurements you took during the interval were sound. When it comes back out of specification, you have a problem that reaches backward: every result taken with it since the last calibration is now in question, and someone has to decide which of those results mattered. Organizations handle that with a documented process, but the trigger is always the as-found reading. Ask for as-found data when sending instruments out; some providers omit it unless requested. Tracking as-found deviations over successive calibrations is also the best way to learn how a particular instrument actually drifts, which is far more informative than any generic assumption.
Choosing an interval that reflects use
Annual calibration is a common default rather than a physical law. The right interval depends on how much the instrument drifts, how consequential its readings are, how roughly it is handled, and what the manufacturer specifies. A bench reference used gently in a stable room may hold specification far longer than a handheld instrument that rides in a truck. Manufacturers usually publish specifications over stated periods, such as ninety days or one year, and an instrument used beyond the period its specification covers has no stated accuracy at all. Where a measurement genuinely matters, shortening the interval is cheaper than an investigation into a year of suspect data. Where it does not, an annual sticker on a continuity tester is mostly ceremony.
What you can do between calibrations
Formal calibration is not the only check available. Keeping a stable reference on the bench, whether a precision resistor, a voltage reference, or simply a known component that never leaves the drawer, lets you verify an instrument in seconds before a measurement you care about. Comparing two instruments against each other does not establish accuracy, but a sudden disagreement between instruments that used to agree is a strong signal that one has changed. Recording those informal checks alongside the date gives you a history. Also treat physical events as triggers: a dropped instrument, a blown input, an exposure to temperature extremes, or a reading that disagrees with expectation should prompt a check regardless of what the sticker says.
What to check before you act
- Request as-found data, not just a pass statement, on every calibration.
- Read the uncertainty stated on the certificate, not only the deviations.
- Confirm the manufacturer’s specification period covers your interval.
- Verify against a stable local reference before measurements that matter.
Common questions
Does a new instrument need calibrating?
It usually ships with a certificate, but check that the certificate is traceable and covers the ranges you will use. A factory statement of conformity is not always the same as a calibration report.
What if an instrument comes back out of specification?
Review what it was used for since its last calibration and decide which results are affected. That backward review is the main reason as-found data exists.
The practical takeaway
Treat calibration as evidence rather than as a ritual. Ask for as-found results, read the stated uncertainty, set the interval from how the instrument is actually used, and keep a local reference for quick checks in between. An instrument whose history you can describe is worth more than one with a recent sticker and no data behind it.
References and further reading
- NFPA 70E: standard for electrical safety in the workplace
- OSHA: 1910.302 scope of electrical design and installation requirements
- CDC/NIOSH: electrical hazards and safe work practices
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
