How often should you calibrate a digital multimeter?
Twelve months is the default answer, but it is rarely the right one for every meter. Here is how to think about calibration intervals for the most common instrument on any bench.
The digital multimeter is the instrument we see more than any other. Handheld meters come in from electrical contractors and maintenance teams by the box; bench 6½- and 8½-digit meters arrive from production test and R&D labs. Nearly every one of them is on a twelve-month interval. Is that right?
Where twelve months comes from
Manufacturers usually publish accuracy specifications for a defined period after calibration, commonly 90 days, one year and sometimes two years. Look at the specification sheet of a typical 6½-digit bench meter and you will see the DC voltage accuracy loosen slightly from the 90-day column to the one-year column. The meter does not suddenly get worse at day 366; the manufacturer is simply guaranteeing a performance envelope for a period they have characterised.
Twelve months is therefore a sensible default: it matches the most commonly used specification column and fits neatly with annual audit cycles.
When twelve months is too long
Consider shortening the interval if:
- The meter is used near the limit of its specification. If your process tolerance is ±0.01 % and the meter’s one-year spec is ±0.0035 %, there is little margin for drift. Use the 90-day specification, and calibrate accordingly.
- It lives a hard life. Handheld meters that travel in vans, get dropped and are used across a wide temperature range drift faster and suffer more damage than bench meters.
- It has failed before. An instrument that has been found out of tolerance at a previous calibration is statistically more likely to do so again.
- The consequences of an error are high. If a meter is used for safety-critical tests or final product acceptance, the cost of a bad measurement outweighs the cost of an extra calibration.
When twelve months might be too short
Equally, a well-cared-for bench meter, used in a temperature-controlled room, with three or four consecutive as-found results comfortably inside tolerance, may be a candidate for a longer interval. That decision should be based on data, not convenience, and documented in your quality system. We will look at formal methods for setting intervals in a later article.
Practical checks between calibrations
Whatever interval you choose, simple intermediate checks catch problems early:
- Measure a known reference, such as a stable voltage reference or a precision resistor, at the start of each week and log the result.
- Compare two meters on the same signal. If they disagree by more than their combined specification, one of them needs attention.
- Check the test leads. A surprising number of “faulty meter” reports turn out to be worn leads with high contact resistance.
Our recommendation
Start at twelve months for most meters. Ask for full numerical results on every certificate, not just pass or fail. After two or three cycles, you will have enough as-found data to see whether each meter is stable, drifting or erratic, and you can adjust the interval with confidence.
