As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A digital multimeter works by converting the quantity you are measuring into a small DC voltage, scaling that voltage with a resistor network, and then digitising it with an analogue-to-digital converter for display. Voltage is measured directly through a high-value input resistor. Current is measured indirectly by passing it through a low-value shunt and reading the voltage across it. Resistance is measured by pushing a known current through the unknown resistor and reading the resulting voltage. Everything else, continuity, diode test, capacitance, temperature, is a variation on those three. The accuracy you get depends far more on the probe contact, the range selected and the burden the meter places on the circuit than on the count rating on the box.
The chain from probe tip to the number on the screen
Every function on a multimeter ends in the same place: a voltage that an analogue-to-digital converter can read. The differences between functions are in what happens before that point.
In voltage mode, the probe tips connect across two points in a circuit. Inside the meter, a large resistor, typically in the region of ten megohms, sits between the input and the converter. The meter draws almost no current, so it behaves like a very high resistance placed in parallel with the part of the circuit you are testing. That is the ideal behaviour for a voltmeter, and it is also the source of the most common bad reading: a high-impedance input will happily display a voltage that exists only because of capacitive coupling or a floating conductor, not because any real source is driving it.
In current mode, the meter inserts a small resistor, called a shunt, in series with the circuit. Current through the shunt produces a voltage across it by Ohm’s law, and the converter reads that voltage. The shunt is deliberately small so that the meter does not disturb the circuit, but it is never zero. That small insertion resistance is called burden voltage, and it is why a current reading is always slightly lower than the current that would flow without the meter in place.
In resistance mode, the meter supplies its own current from an internal source, passes it through the unknown resistance, and measures the voltage developed across it. This is why resistance must be measured on a de-energised circuit. If any external voltage is present, it adds to or subtracts from the meter’s own test voltage and the reading becomes meaningless, or the meter is damaged.
Continuity, diode test, capacitance and temperature are all built on the same core. Continuity is a resistance measurement with a threshold and a beeper. Diode test is a voltage measurement while a small current is forced through the junction. Capacitance is measured by timing how long the unknown capacitor takes to charge through a known resistor. Temperature is a resistance measurement of a thermocouple or resistive sensor, converted using a stored curve.
Why the count rating matters less than the range you selected
A 6000 count meter can display 5999 before it changes range. That figure describes the resolution at the top of a range, not the accuracy of the reading. Accuracy is usually published separately as a percentage of reading plus a number of counts, and the counts term dominates when the reading is small.
Consider a meter specified at plus or minus 0.5 percent of reading plus 2 counts. On the 6 volt range, a 5 volt reading has an uncertainty of about 0.025 volts from the percentage term plus 2 counts, which on that range is roughly 0.002 volts. On the 60 volt range, the same 5 volt reading has the same percentage term but the counts term is ten times larger. Selecting the lowest range that does not overload is not a fussy habit; it is the difference between a reading that is useful and one that is mostly noise.
Autoranging meters do this for you, but they do it on their own schedule. A slowly changing signal near a range boundary can make the display flicker between two ranges, and each range has a different counts contribution. If you need a stable number, manual range selection is often the better choice.
The update rate matters for a different reason. A meter that updates four times a second shows you four samples a second. A fast-changing signal, a motor start, a relay chatter, will be under-sampled, and the display may never show the peak. The analog bar graph on better meters updates faster than the digits, which is why it is useful for peaking and nulling adjustments even though it is less precise.
True RMS, averaging and the waveform you cannot see
Most inexpensive meters measure the average of the rectified waveform and then scale that average to display what the RMS value would be for a pure sine wave. On a clean sine wave this is correct. On a distorted waveform, a phase-controlled dimmer, a variable frequency drive, a switched-mode power supply, the average and the RMS are not related by that fixed factor, and the meter reads low. True RMS meters compute the heating equivalent of the waveform directly, so the reading stays correct as the shape changes. The distinction is explained in more detail in what true RMS means on a multimeter.
True RMS has its own limit. Most true RMS meters are specified only up to a certain crest factor, the ratio of peak to RMS. A waveform with a high crest factor, such as a short pulse, can still read low even on a true RMS instrument. The listing rarely publishes the crest factor limit, and where it does not, treat readings on highly pulsed loads with caution.
This is where a meter built for a specific trade earns its place. The Fluke 116 HVAC multimeter adds a built-in thermometer covering minus 40 to 400 degrees Celsius and a microamp range for testing flame sensors, because the HVAC technician needs a temperature reading and a microamp reading in the same visit as the voltage checks. The microamp range matters because a flame rectification signal is on the order of a few microamps, and a meter without that range cannot resolve it.
The shunt, the clamp and the reason current is the awkward measurement
Current measurement is the one function that requires you to break the circuit or accept a compromise. A series shunt meter must be inserted in line, which means disconnecting a terminal and reconnecting through the meter. That is slow, and on live equipment it is hazardous.
A clamp meter avoids the break by measuring the magnetic field around a conductor. The clamp jaws form a magnetic circuit, and a Hall sensor or a current transformer converts the field into a voltage. Clamp meters therefore measure current without contact, but they only work on a single conductor. Clamping around a cable that carries both line and neutral reads the difference between the two currents, which for a healthy load is close to zero. This is the most common mistake with a clamp, and it is covered in multimeter versus clamp meter.
The POROMETISTO CM06 clamp multimeter illustrates two functions that exist because current is awkward. The inrush function captures a short window, about 100 milliseconds according to the listing, to catch the startup surge of a motor. The listing states that this function applies to AC only. The LoZ mode places a low impedance across the input to drain stray voltages that would otherwise give a phantom reading on a high-impedance input. Both are answers to specific failure modes, not general improvements.
For very high voltage work, the measurement chain has to be built for the voltage, not just protected against it. The KPS DMM4500PV solar multimeter is listed as measuring DC voltage up to 2000 volts and AC up to 1500 volts, with Bluetooth logging to a phone app. The voltage rating is a property of the input divider and the isolation, and it is the first specification to check before touching a photovoltaic string.
Where the meter stops being the right instrument
A multimeter measures electrical quantities. It does not measure insulation resistance, it does not measure earth loop impedance, and it does not verify that a circuit is dead in the sense a safety procedure requires. A voltage reading of zero on a high-impedance meter can mean the circuit is dead, or it can mean the meter is not connected to the source. The difference between a voltage tester and a multimeter is exactly this: a tester is designed to prove presence or absence under load, while a meter is designed to quantify. That distinction is set out in voltage tester versus multimeter.
A multimeter also does not replace a dedicated insulation tester for cable testing, because the test voltage required is far above what a multimeter can supply. It does not replace a dedicated earth resistance tester. And it does not replace a power quality analyser, because it samples too slowly and cannot record harmonics or transients over time.
Where a meter does excel is in the diagnostic work that precedes all of those: checking that a supply is present, that a fuse is intact, that a contactor coil is being energised, that a sensor is returning a plausible resistance. The Fluke 177 true RMS multimeter is a useful example of the display side of that work. The listing states 6000 counts, a digital update four times a second and a 33 segment analog bar graph updating forty times a second, plus a hold function and an auto hold that waits for a stable input before capturing. The fast bar graph is the part that matters when you are watching a signal change while adjusting something.
The maintenance nobody mentions: probes, fuses and the calibration date
The meter is usually the most reliable part of the measurement. The probes are the part that fails. A probe with a broken strand inside the insulation will read correctly until the moment it is flexed, and then it will read open or intermittently. Probe tips oxidise, and a dull tip on a terminal screw produces a contact resistance that adds to every reading. Replacing probes is cheap and it is the single most effective maintenance action on a multimeter.
The current input fuse is the second item. A meter that has been accidentally connected across a supply while in current mode will have blown its fuse, and many meters will still display a reading on the voltage ranges, which hides the fault. Check the fuse if any current reading looks wrong.
Calibration is the third. A meter drifts, and the drift is small but it is real. For work where the number has to be defensible, a calibration certificate with a stated uncertainty is the only way to know the reading is traceable. For general diagnostic work, comparing the meter against a known source, a fresh battery and a stable reference, is enough to catch a gross error. The procedure for checking your own readings is covered in how to use a multimeter.
| Function | What the meter actually does | Typical failure mode |
|---|---|---|
| DC voltage | High-impedance divider into the converter | Phantom voltage on floating conductors |
| AC voltage | Rectifier or true RMS converter, then divider | Low reading on distorted waveforms if not true RMS |
| Current | Shunt in series, voltage across shunt | Burden voltage; blown fuse after a misconnection |
| Resistance | Internal current source through the unknown | Wrong reading if the circuit is not de-energised |
| Continuity | Resistance with a threshold and beeper | Beeps through a semiconductor junction |
| Temperature | Sensor resistance or thermocouple voltage | Wrong probe type selected |
None of the products referenced here publishes a calibration interval or an accuracy figure in the supplied bullets. Where you need those, they are in the manufacturer’s datasheet, not in the marketing copy. Not stated in the listing bullets is the honest answer for any figure not given above.
What the number on the display is really telling you
A multimeter is a voltage measuring instrument with a set of adapters around it. Once you see the shunt, the divider and the internal current source as the three mechanisms behind every function, the failure modes become predictable: a high-impedance input that believes a stray field, a shunt that drops voltage it should not, a resistance measurement contaminated by an external source. Choose a meter for the ranges and functions your work actually needs, check the probes before you blame the meter, and treat any reading that matters as provisional until you have confirmed it a second way.
Frequently Asked Questions
Does a multimeter measure current directly?
No. It measures the voltage across a small internal resistor, the shunt, and calculates the current from Ohm's law. That is why the meter has to be placed in series with the circuit, and why the reading is slightly affected by the shunt resistance.
Why does my meter read a voltage on a wire that is not connected to anything?
The input impedance is very high, so the meter can display a voltage coupled capacitively from a nearby live conductor. A low-impedance mode, often labelled LoZ, drains that charge and shows the true reading. If your meter has no LoZ mode, a small load placed across the probes will do the same job.
Is a 6000 count meter more accurate than a 2000 count meter?
Not necessarily. Counts describe resolution at the top of a range. Accuracy is published separately as a percentage plus counts, and a lower-count meter with a tighter percentage specification can be more accurate on the range you are using.
Can I use a multimeter to test a solar string?
Only if the meter's DC voltage rating exceeds the string's open-circuit voltage with margin. A standard 600 volt meter is not suitable for a high-voltage string. The KPS DMM4500PV is listed at 2000 volts DC, which is the kind of rating that work requires, but check the actual string voltage before connecting anything.
What does true RMS actually change?
It changes the reading on non-sinusoidal waveforms. An averaging meter reads low on a distorted wave because it scales the rectified average as if the wave were a sine. A true RMS meter computes the heating equivalent directly, so the number stays correct as the shape changes.
Does a clamp meter work on a two-core cable?
No. A clamp measures the magnetic field around a single conductor. Around a cable carrying line and neutral, the fields cancel and the clamp reads near zero even when current is flowing. You must clamp one conductor at a time.