MEASUREMENT, WITH CONTEXT.Source-led publishing · UK English
Meter & ProbeTHE MEASUREMENT FIELD NOTES

READING THE SPECIFICATION · 6 MIN READ

More counts. More accuracy?

Decode multimeter resolution and accuracy specifications, then calculate a stated error limit for a particular range and reading.

THE SHORT ANSWER

Counts describe display capacity; they do not tell you how close a reading is to the true value. To compare meters at a particular voltage or current, use the resolution and accuracy specification for the selected range, including the digits term and all stated conditions. A finer display can still have a larger specified error.

Three numbers that do different jobs

A 6,000-count instrument can typically display values up to 5,999 on a range. That figure describes the display, not a universal error limit. The range determines what the last displayed digit represents. On an illustrative nominal 6 V range displaying up to 5.999 V, the smallest displayed step is 0.001 V. Another range on the same meter may have a different step and a different specification.

TermWhat it tells youWhat it does not establish
ResolutionThe smallest displayed increment on this rangeThe smallest change you can measure reliably
Accuracy specificationA stated error limit under specified conditionsThe actual error of your individual instrument
RepeatabilityHow closely repeated readings agreeAgreement with a traceable reference

HKJ’s tolerances explanation [1] separates the percentage and digits terms. A stable final digit can look reassuring while the result remains offset from the reference value.

Turn the specification into an error limit

A common format is ±(percentage of reading + a number of digits). Here, “digits” means counts of the least significant displayed digit on the chosen range. It does not mean the total number of characters on the screen.

E = |reading| × percentage / 100 + digits × resolution

For a hypothetical current range with 1 mA resolution and a specification of ±(2% + 8 digits), a 20 mA reading gives E = 0.4 mA + 8 mA = 8.4 mA. The simple specified interval is therefore 11.6–28.4 mA. The ±8.4 mA error limit is 42% of the displayed reading, despite the 1 mA display step.

This is an arithmetic example, not a UT210E specification or a measurement of any product. Use the exact manual for the model, revision, function and range you are considering.

Try your range specification

SPECIFICATION → ERROR LIMITCalculated example
Example: 20 mA ± 8.4 mA → 11.6 to 28.4 mA (42% of reading).

Enter the specification for the actual function and range. This is a stated error-limit calculation, not a confidence interval or calibration uncertainty. Extra specification terms are not included.

Check the conditions before comparing meters

Look for the temperature interval, time since calibration, warm-up requirement and any temperature coefficient. For AC measurements, frequency, minimum amplitude and waveform restrictions matter too. Additional terms outside the specified conditions must be evaluated separately; the calculator does not invent them.

Compare both meters at the same intended measurement point. At 10 mV, range selection and the digits term may dominate. At 12 V, a different range may apply. Quoting a meter’s best DC percentage from its headline specification does not establish its performance at either point.

Auto-ranging can change the step size part-way through a measurement. Record the actual range rather than assuming that the finest advertised resolution applies to every reading.

A specification is not a calibration certificate

The calculated interval is neither a statistical confidence interval nor a complete measurement uncertainty budget. A calibration result relates a particular instrument to references, with stated uncertainty and conditions. Noise, leads, source stability and the measuring circuit can add further limitations.

Two meters that agree may share an error. To investigate a discrepancy, keep the source, connections, range and timing consistent, and use a suitable reference. Our measurement record captures those conditions without claiming that a comparison constitutes traceable calibration.

Sources & scope

Sources checked on 22 September 2026. Numbered references above point to the original publishers below. Their sample measurements are not Meter & Probe measurements; manufacturer material is identified separately.

  1. HKJ — Digital multimeter tolerances ↗Independent measurement reference
  2. HKJ — How digital multimeters are tested ↗Independent test methodology

No product in this guide has been bench-tested or certified by this publication. Evidence policy · Corrections

Keep the conditions with the reading.

Use the blank CSV to record the instrument, range, reference, environment and result.

Download measurement record ↓

KEEP READING