Why Does My Multimeter Show the Wrong Voltage?

What counts as a wrong reading

Before blaming the meter, decide what you expected and why. A 9 V battery that reads 8.4 V is probably just partly used. A US outlet that reads 118 V or 123 V is normal, because utility voltage is allowed to vary around the 120 V nominal. A reading only counts as wrong when it disagrees with the source by more than the meter's stated accuracy plus normal variation in the circuit, or when it is obviously impossible: 0.3 V on an outlet that powers a lamp, or 47 V on a 5 V USB port.

Wrong readings fall into two groups. Setup errors come from the dial, the jacks, the leads or the battery, and they are fixed in seconds. Measurement effects are cases where the meter is accurately reporting what is at its input, but that is not the value you wanted: the waveform is not a clean sine, the meter's own resistance changes the circuit, a floating conductor is picking up a coupled voltage, or the black lead is on a different reference than you think.

Causes at a glance

1Check the setup first

Function. AC and DC are separate positions on most dials, or one position with a SELECT button that toggles between them. Measure a car battery on V AC and many meters show close to zero; measure a wall outlet on V DC and you get a small number that wanders. Many dials also have a separate mV position for small signals. On mV, a 1.5 V cell either shows OL or forces the meter into a range that confuses the reading, so check the unit printed beside the digits, not just the dial. If you switch between AC and DC circuits all day, a meter with automatic AC/DC selection removes the step: Fluke lists AutoVolt as one of the features the 117 adds over the 115.

Range. On a manual-range meter, a range set far above the signal throws away resolution: 0.4 V on the 600 V range may display as 0.0 or 0.4 with no further digits. A range set below the signal shows OL. Auto-ranging meters handle this, but they need a moment to settle, and a reading taken while the meter is still stepping through ranges can be wrong.

Jacks. Voltage is measured with the red lead in the V (often V/Ω) jack. The A and mA jacks connect the leads across a low-value shunt resistor, so a lead left there after a current test reads almost nothing and, on a source that can deliver current, blows the meter's fuse or worse. Meters with lit or guarded jacks, such as Klein's Lead-Alert models, make this mistake easier to catch.

Battery and leads. A weak battery can shift readings before the low-battery symbol appears on some meters, so a fresh one is the cheapest test there is. Then short the probe tips together on the resistance range: a reading that jumps or stays high points to a damaged lead or a loose plug, which can also make voltage readings read low on high-impedance sources. If the number will not settle at all, that is a different problem, covered in unstable multimeter readings.

2Compare against a known source

Once the setup is right, measure something whose voltage you can trust. Good choices are a voltage reference module or precision reference IC, a regulated bench supply checked by a second meter, or a fresh alkaline AA cell (typically around 1.5 to 1.6 V when new, so it is a sanity check rather than a calibration standard). For AC, a known-good outlet is a reasonable reference for spotting gross errors, since the utility keeps it near nominal.

If the meter reads the reference correctly, the meter is fine and the problem is in the circuit or in how you are measuring it. If it reads the reference wrong on DC as well as AC, with a fresh battery and good leads, suspect the meter itself. Our guide to measuring a voltage reference covers how to do this carefully.

3Readings that are real but unexpected

True RMS vs average-responding. Most budget and some name-brand meters measure the average of the rectified AC waveform and scale it as if the signal were a pure sine wave. On a clean sine that gives the right answer. On a chopped or distorted waveform, from a dimmer, a variable-speed drive, a modified-sine inverter or a switching supply, the scaling no longer fits and the reading can be noticeably high or low depending on the waveform shape. A True RMS meter computes the effective value directly, so it stays meaningful on those signals within its bandwidth and crest-factor limits. Fluke's own product pages state that the 106 and 107 are not True RMS, while the 115, 117 and 87V are, which is a useful example of the split even within one brand. The AC vs DC voltage guide explains what an RMS figure represents.

Ghost voltage. A wire that is switched off but runs alongside live conductors can show a convincing AC voltage on a high-impedance meter, because the input barely loads the tiny coupled current. That is covered in depth in why a multimeter shows voltage on a disconnected circuit.

Reference point. A voltage is always between two points. Clip COM to chassis in a car and probe a sensor output, and you include any voltage dropped in the sensor's ground path. Probe a split-rail circuit from the wrong rail and you see the sum of two supplies. If the size of the number looks right but the sign is reversed, see why a multimeter shows negative voltage.

4When the meter loads the circuit

On its DC voltage function a typical digital multimeter presents an input impedance of around 10 MΩ. Across a battery or a regulated rail that is negligible. Across a node fed through resistors in the megohm range, such as a high-value divider, a pH or piezo sensor, or a reference with a large output resistor, the meter becomes part of the circuit and drags the voltage down. The diagram shows the effect with two 1 MΩ resistors: the meter's 10 MΩ in parallel with the lower resistor makes about 0.91 MΩ, so the midpoint sits nearer 4.76 V than 5 V.

The tell is that the error grows as the circuit's resistance grows, and that a meter with a different input impedance gives a different answer. Check the input impedance in your meter's manual (some differ between AC and DC, and LoZ modes are deliberately low), measure at a buffered or lower-impedance point if the design has one, or calculate the loading and correct for it. For more on measuring across individual parts, see how to measure voltage across a resistor.

5Accuracy and calibration

Every meter's accuracy is written as a percentage of the reading plus a number of counts on the last digit, and it differs by function and range. A spec of ±(0.5% + 2) on a 6.000 V range allows a 5.000 V source to read anywhere from about 4.973 to 5.027 V, so two good meters can disagree in the last digits without either being wrong. AC accuracy is usually looser than DC and is specified only over a frequency band. The multimeter accuracy guide walks through reading these specs.

A meter that is consistently outside its own spec on a trusted reference, after the battery and leads have been ruled out, has drifted or been damaged, often by an overload on the wrong function. Inexpensive meters are usually replaced rather than calibrated. Higher-end meters can be sent to an accredited lab, and anyone whose work depends on traceable readings normally does that on a fixed schedule.

Why does my multimeter show different voltage readings?

When a multimeter gives different voltage readings on the same circuit, whether two meters disagree or one meter disagrees with a label or a second test point, the cause is usually one of four things: the meters measure AC differently, their input impedances load the circuit differently, both are within their accuracy specs but at opposite ends, or one of them is reading a coupled ghost voltage. Each has a recognizable size, so the gap itself is a clue. (A single reading that keeps changing is a separate problem, covered in unstable multimeter readings.)

True RMS against average-responding. An average-responding meter multiplies the rectified average by about 1.11, the ratio that holds for a pure sine. Feed it a square wave, where the average and the RMS are equal, and it reads about 11% high: a 120 V RMS square-wave output shows as roughly 133 V. On peaky waveforms, such as the current drawn by electronic power supplies, the same scaling tends to read low instead. If one meter is True RMS and the other is not, a gap on a dimmer, inverter or drive output says nothing is broken. Our True RMS multimeter roundup lists affordable meters that measure distorted waveforms properly.

Different input impedance. Go back to the 1 MΩ / 1 MΩ divider in step 4. The same node reads differently on every meter because each one adds its own resistance in parallel with the lower leg. The chart shows the arithmetic for four common input impedances. A classic analog meter rated 20 kΩ per volt presents 200 kΩ on its 10 V range, so it reads less than a third of the true value; a meter in LoZ mode, designed to load the circuit on purpose, reads lower still.

Accuracy specs that overlap. Suppose meter A is specified at ±(0.5% + 2 counts) and meter B at ±(1.0% + 3 counts), both reading a 12.00 V battery on a range with 0.01 V resolution. A may show anything from 11.92 to 12.08 V (0.5% of 12 V is 0.06 V, plus 0.02 V), and B anything from 11.85 to 12.15 V (0.12 V plus 0.03 V). They can disagree by 0.23 V while both are working perfectly. Look up the spec for the exact function and range before deciding which meter is wrong.

Ghost voltage on one meter only. If one meter shows tens of volts on an "off" wire and another shows nearly zero, the second one is probably a low-impedance meter or in LoZ mode, which drains the tiny coupled charge. The detail is in voltage on a disconnected circuit; the practical rule is that the wire stays treated as live until it is proven dead.

Why is my multimeter reading incorrect DC voltage?

An inaccurate DC voltage reading usually traces to the meter's own battery, the wrong range or unit, AC ripple riding on the DC, a black lead on a different reference than you think, or a stored REL offset. Calibration drift is real but it is the least common cause, so check it last. The table gives the typical size of each error, which is often enough to identify it.

AC ripple on DC. The DC function reports the average of whatever is at its input. A simple battery charger, or any rectifier without a large smoothing capacitor, delivers humps rather than a flat line. From a 12 V RMS winding the peaks reach about 12 × 1.414 ≈ 17.0 V, but the average of a full-wave rectified sine is 0.637 of the peak, so the meter shows about 10.8 V (ideal diodes; real bridge rectifiers lose roughly another volt or more). Neither number is wrong; they describe different things. Switching to V AC on the same points reveals how much ripple is there, and a scope shows its shape.

Reference point. In a car, a sensor's 5 V reference measured to a chassis bolt includes any voltage dropped in the ground path between that bolt and the module. A corroded strap dropping 0.3 V makes a perfect 5.00 V supply read 4.70 V. Measure to the module's own ground pin, or to battery negative, and compare; the difference is the ground drop, which is worth fixing on its own. Our guide to testing a car ground wire shows how to find it. On isolated supplies there may be no shared reference at all, which measuring on a floating ground explains.

Battery and calibration. If the DC error is the same proportion on every source, swap the meter's battery and measure a reference again. If it is still outside the published spec for that range, the meter has drifted. Pressing REL by accident gives a different signature: a fixed offset, such as every reading being 0.50 V low, regardless of what you measure.

Meters that help

No meter prevents every mistake, but some features remove whole rows from the table above: True RMS for distorted AC, automatic AC/DC selection, and LoZ for ghost voltages. These picks come from our verified multimeter list.

Meters with True RMS and mistake-proofing features

Fluke 117

True RMS, CAT III 600V, with AutoVolt AC/DC selection, LoZ and NCV. Hands-on Tested.

Check price on AmazonRead the review →

Fluke 87V

True RMS with a 19,999-count high-resolution mode and CAT IV 600V. Hands-on Tested.

Check price on AmazonRead the review →

KAIWEETS HT118A

Budget True RMS, 6,000-count, with NCV and Min/Max. Maker's page and manual disagree on max voltage.

Check price on AmazonRead the review →

As an Amazon Associate I earn from qualifying purchases. Prices and availability are shown on Amazon.