Compliance and audits
Multimeter calibration: procedure, test points and how often to calibrate
Multimeter calibration applies known values from a traceable multifunction calibrator at set points on each range of a digital multimeter (DC and AC voltage, DC and AC current and resistance) and compares each reading with the maker's specification, which is commonly written as a percentage of the reading plus a number of counts. No standard fixes one interval for every meter: you set it from use and the meter's own as-found history, starting from the period the maker's specification covers.
On this page
What a multimeter calibration checks
A calibration applies known values from a reference standard and records what the multimeter reads at each point. The error is the reading minus the applied value. It is recorded with the uncertainty of the calibration. Comparing each error with the maker's specification and recording a pass or fail is a verification, and a certificate can carry both.
EURAMET Calibration Guide No. 15 (cg-15, version 3.0, 2015), from the association of Europe's national metrology institutes, sets out a method for the job. It covers DC voltage, AC voltage, DC current, AC current and resistance on digital multimeters. It leaves out panel meters, instruments built mainly for power, energy or impedance, and measurements above 1 MHz. It supplements the maker's calibration instructions rather than replacing them, so follow the maker where it asks for more.
Other functions, such as capacitance and frequency, get a smaller set of points. An accessory the meter normally measures through, such as a current shunt, is calibrated with the meter in the same setup.
How multimeter specifications work
Makers commonly state a multimeter's accuracy as ±(percentage of reading + counts). The percentage scales with the reading. The counts are a fixed number of steps of the last displayed digit on that range. That fixed part is why a meter is less accurate, relative to the reading, near the bottom of a range. Some makers write digits instead of counts. Others give a percentage of the range in place of the counts.
Take a meter specified at ±(0.5% of reading + 2 counts) on a 6 V range that reads to 0.001 V. At a 5.000 V test point the limit is 0.5% of 5 V, which is 0.025 V, plus 2 counts of 0.001 V. That makes ±0.027 V, so a reading from 4.973 V to 5.027 V is inside the specification. At 0.600 V on the same range the limit is ±0.005 V, or about 0.8% of the reading.
A specification commonly comes with conditions, such as a temperature range, a warm-up time and a period since calibration. When EURAMET cg-15 weighs the calibration uncertainty against the specification, it uses the one-year specification, because that is the one users generally rely on while the calibration is valid. Use the specification for the period that matches your interval.
Resolution is often given as a number of digits. EURAMET cg-15 splits multimeters into two groups at 4¾ digits, or 39,999 counts. Lower-accuracy working meters sit at or below it, and higher-resolution meters above it. The two groups get different sets of test points.
Which test points to use
Test points are set as a percentage of each range's full scale. EURAMET cg-15 uses 10% for the start of a range and 90% for its top. It does not recommend points below 10%, because the next range down measures them more accurately. The 90% point may sit anywhere from 50% to 99% of full scale, and from 30% for resistance.
Each range is calibrated at least near its top. Linearity is checked on one DC and one AC range with at least five points, the AC points at a fixed frequency, usually 1 kHz. For a meter of up to 4¾ digits the minimum set in cg-15 looks like this.
| Function | On each range | Extra points |
|---|---|---|
| DC voltage | +10% (zero on the smallest range) and ±90% | On one intermediate range: ±10%, +50% and ±90%, with at least five points spread evenly |
| DC current | +90%, with zero on the smallest range | On one intermediate range: +10% and ±90%, plus 50% on ranges above 1 A |
| Resistance | +10% (zero on the smallest range) and +90% | None |
| AC voltage | 90% at three or more frequencies between 20 Hz and 100 kHz | On one intermediate range: 10%, 50% and 90% at 50 Hz and 1 kHz. On the smallest range: 10% at one frequency |
| AC current | 90% at three or more frequencies between 20 Hz and 5 kHz | On the smallest range: 10% at one frequency |
| Other functions, such as capacitance and frequency | 10% (zero on the smallest range) and 90% | None |
Equipment and conditions
The usual reference is a multifunction calibrator, which sources DC and AC voltage, current and resistance from one instrument. Its own calibration has to be current and traceable, with an uncertainty small enough for the meter under test. A 2001 NIST report on its Josephson volt standard notes that high-end multimeters can be as accurate, and resolve as finely, as the multifunction calibrators made to calibrate them, so a high-end laboratory meter (up to 8½ digits) may need better standards than a multifunction calibrator offers.
Traceability runs from the calibrator's certificate, through the labs above it, to a national metrology institute. There the volt is realized with the Josephson effect and the ohm with the quantum Hall effect, as the SI Brochure's practical realization of the electrical units describes.
- A multifunction calibrator, or separate voltage, current and resistance standards, with current certificates.
- Leads and cables in good condition. For resistance, note whether the meter measured in two-wire or four-wire mode. In two-wire mode, null the lead resistance by shorting the leads at the resistor end.
- A low thermal EMF short for zeroing the DC voltage ranges.
- A stable room, with the temperature and humidity recorded. Give the meter the stabilization time its maker specifies, powered on if it runs from the mains.
- The maker's calibration manual, which sets the adjustment method and any points the maker requires.
As-found readings, adjustment and as-left readings
EURAMET cg-15 describes three possible sequences. The standard one for a meter that can be adjusted runs a calibration before adjustment, the adjustment, and a calibration after it. The first gives the as-found readings, which show how the meter behaved since its last calibration. The last gives the as-left readings. ISO/IEC 17025:2017 clause 7.8.4 asks a calibration certificate to report the results before and after any adjustment or repair where they are available.
A meter does not have to be adjusted at each calibration. cg-15 allows calibration without adjustment when the readings sit well inside the specification. For a lower-accuracy meter with manual adjustment, its example limit is 70% of the one-year specification at all points, and 50% at points where the meter is adjusted. For more accurate meters the example limit is 50% at all points.
Adjustment follows the maker's method. Depending on the meter, that means adjusting internal components or running a firmware routine.
If an as-found reading is outside the specification, ISO 9001 clause 7.1.5.2 asks you to determine whether earlier results were adversely affected and to act where needed. Look at what the meter was used for since its last good calibration, such as acceptance tests, equipment settings or checks of other instruments. Judge whether an error of that size and sign could have changed a decision, and record the outcome.
TUR and the 4:1 ratio
The test uncertainty ratio (TUR) compares the limit being checked with the expanded uncertainty of the calibration at that point. For a symmetric limit of ±T and an expanded uncertainty U at about 95 percent coverage, TUR = T / U. EURAMET cg-15 says a ratio of 4:1 between the specification and the calibration uncertainty is usually sufficient, taking the one-year specification.
In the example above the limit at 5 V is ±0.027 V. A 4:1 ratio there needs an expanded uncertainty of no more than about 0.0068 V. cg-15 accepts that 4:1 is not always possible. On low-resolution meters the last digit dominates both the specification and the uncertainty. On the most accurate meters the standards may fall short on some ranges. It asks labs to keep the points below 4:1 to a minimum.
Resolution adds uncertainty because each reading is rounded to its last digit. The GUM (JCGM 100:2008, F.2.2.1) treats that rounding as a rectangular distribution one step wide. Its standard uncertainty is the step divided by √12, about 0.29 times the step. cg-15 asks the certificate to make clear that its uncertainties include the meter's resolution and short-term instability.
Whether a reading passes also depends on the decision rule. ISO/IEC 17025:2017 clause 7.8.6 requires a lab that states conformity to document the decision rule it used, and cg-15 points to ILAC-G8 for judging and reporting compliance. Ask your lab whether it uses simple acceptance or a guard band.
How often to calibrate a multimeter
No standard sets one interval for all multimeters. ISO 9001 asks for calibration or verification at specified intervals, or before use. ILAC-G24 / OIML D 10 (2022 edition) says there is no single best practice. It lists factors to weigh, among them the uncertainty you need, the maker's recommendation, the tendency to drift, how much and how hard the meter is used, its environment and the results of intermediate checks.
A practical starting point is the period the maker's accuracy specification covers. Then adjust from the meter's own as-found history. Meters that come back well inside their limits support a longer interval. A failure, or a drift toward the limit, supports a shorter one. ILAC-G24 advises calibrating new equipment more often at first to see its trend.
Between calibrations, check the meter against something stable and known. A voltage reference, a calibrator or a fixed resistor that you measured right after the last calibration will do. ISO/IEC 17025:2017 clause 6.4.10 asks labs to carry out intermediate checks where they are needed to keep confidence in the equipment. After a drop, an overload or a blown fuse, check the meter before it is used again.
What the certificate and record should show
EURAMET cg-15 lists what a multimeter certificate needs beyond the usual identification: the method, evidence of traceability, the conditions, the results with their units and the uncertainty. In practice the record should show the following.
- The meter's ID, make, model and serial number.
- The procedure used, and whether a self-test, self-calibration, adjustment or repair was done.
- The reference standards used, with their certificate numbers and due dates.
- The temperature and humidity, and how long the meter stabilized before the readings.
- The meter's settings, such as two-wire or four-wire resistance and the input used.
- For each point: the range, the applied value (with the frequency for AC), the reading as found and as left, the error and the uncertainty.
- The specification and the decision rule behind any pass or fail statement.
- The date, the person or lab, and the next due date.
How to calibrate a multimeter, step by step
- Check that the calibrator and any other standards have current certificates and can reach about 4:1 against the meter's one-year specification.
- Let the meter settle in the room and power it for the warm-up time its maker gives. Record the temperature and humidity.
- Inspect the meter, its leads and its input terminals. Run its self-test, and its self-calibration if it has one.
- Zero the DC ranges where the meter allows it: a low thermal EMF short for voltage, an open input for current and shorted terminals for resistance.
- Apply each test point from the calibrator and record the as-found reading. Cover each range of each function, with the extra linearity points on one DC and one AC range.
- Compare each error with the maker's specification for that range and point, using your documented decision rule.
- If the meter needs adjustment, adjust it by the maker's method. Then repeat the test points and record the as-left readings.
- Record the results, the standards used, the uncertainty and the next due date, and label the meter with its status.
How Axiospec helps with multimeter records
Axiospec is calibration management software. Each multimeter gets a record with its calibrations, its certificates and its next due date. A calibration can carry test points, each with its nominal value, its tolerance and the as-found and as-left readings. A tolerance can be an absolute value or a percentage of the nominal or of the span, so a limit in percent of reading plus counts goes in as the absolute value worked out for that point.
Register your calibrator as a reference standard and pick it on each calibration. Its own record then lists the calibrations it was used on. When a calibration is logged as Fail, the meter goes into quarantine and the record asks for the out-of-tolerance impact. Axiospec does not calibrate the meter. On an in-house calibration, a final reading outside its tolerance sets the result to Fail, and recording a pass anyway takes a written reason. The tolerance, the decision rule and the interval stay with you or your lab.
It is free for up to 50 active instruments. A new workspace starts with a 45-day trial of Professional, with room for 1,500 active instruments. We load your existing records for you at no charge on every plan.
Common questions
- How often should a multimeter be calibrated?
- No standard sets one interval. Start from the period the maker's accuracy specification covers, then lengthen or shorten it from the meter's as-found results, its use and its environment, as ILAC-G24 recommends. Check it against a known source between calibrations and after any drop or overload.
- How do you calibrate a digital multimeter?
- Let it warm up, zero the DC ranges and apply known values from a traceable multifunction calibrator on each range of each function, typically at 10% and 90% of full scale, with more points on one DC and one AC range. Record each reading as found and compare its error with the maker's specification. If the meter needs adjustment, adjust it and repeat the points as left.
- What does ±(% of reading + counts) mean on a multimeter?
- It is the meter's accuracy limit. The percentage applies to the reading, and the counts are a fixed number of steps of the last digit on that range. A meter specified at ±(0.5% + 2 counts) and reading 5.000 V on a range that resolves 0.001 V has a limit of ±0.027 V.
- What equipment do you need to calibrate a multimeter?
- A multifunction calibrator, or separate voltage, current and resistance standards, with a current traceable calibration. At each point its uncertainty should be about a quarter of the meter's specification or smaller, the 4:1 ratio EURAMET cg-15 calls usually sufficient. You also need good leads, a low thermal EMF short, a stable room and the maker's calibration manual.
- Why is a multimeter calibrated at 10% and 90% of each range?
- EURAMET cg-15 uses 10% of full scale for the start of a range and 90% for its top, and calibrates each range near its top. It does not recommend points below 10%, because the next range down measures those values more accurately.
- Can I calibrate my own multimeter?
- Yes, if you have a calibrator with a current traceable calibration and enough accuracy, a written procedure and trained people. ISO 9001 does not require an outside or accredited lab for it. Customers can still ask for accredited certificates, and adjustment needs the maker's method.
Put it into practice
Got a gage list? Send it over and we load it for you, usually in a couple of business days. Free on every plan. Then log every calibration to a tamper-evident audit trail and produce records on demand.
Axiospec is a documentation and workflow tool. It helps you keep clean, traceable, audit-ready records. Certification depends on your own processes, scope and assessor.
Take the audit-prep checklist with you
First the checklist, with each item phrased the way an auditor asks it. Then four short emails with practical tips. You can unsubscribe in one click.
First we email you a link to confirm your address. Nothing else comes until you confirm, and you can unsubscribe from any email after that. See our Privacy Policy.