What Loading Does When You Connect an Instrument

Why this matters

A meter across two points is a resistor across two points. A meter in series is a resistor in series. Neither one is a neutral observer, and the amount they change what they are reading is not fixed, it depends entirely on what they are connected to. This is why a tech can stand at a control circuit, read full voltage on the display, and be looking at a circuit that cannot deliver enough energy to pull in a contactor. Nothing is wrong with the meter. The meter is reporting exactly what it sees while drawing almost no current, and the fault only exists when current flows.

Techs who learn this one relationship stop chasing intermittents that were never intermittent, and stop condemning boards on the strength of a voltage that was never there.

Before you put an instrument into a circuit

Any of this on a live circuit sits behind the energized-work gate at 29 CFR 1910.333(a)(1), which requires de-energizing before work on or near live parts unless that is infeasible and treats a test that can only be performed energized as one of those cases. Use the electrical protective equipment 29 CFR 1910.335(a) requires for the exposure, the arc-flash boundary and PPE your employer's electrical safety program sets under NFPA 70E-2021 in the edition it adopts, and keep the free hand clear of the enclosure.

Inserting an ammeter in series means opening the circuit. That is no longer a test that can only be performed energized: de-energize, apply electrical lockout under 29 CFR 1910.333(b)(2), or 29 CFR 1926.417 on construction, prove dead with the live-dead-live sequence in NFPA 70E-2021, 120.5, break the circuit and land the meter, then re-energize with hands clear.

A current input placed across a voltage source is a bolted fault through your instrument. It is the single most common way a meter is destroyed and a hand is burned. Confirm the leads are in the voltage jacks and the function selector agrees with them before the tips go anywhere, and confirm the current input still carries the manufacturer's specified high interrupting-rating fuse, which a sibling article covers (see References).

Two ways an instrument loads a circuit

Across (shunt). A voltmeter is a large resistance placed in parallel with the thing being measured, so it steals a small current that used to go somewhere else. The larger its input impedance, the less it steals. A typical hand-held meter's voltage input is on the order of 10 megohms.

Through (burden). An ammeter is a small resistance placed in series, so it drops a small voltage that used to appear across the load. That drop is called burden voltage, and it is the shunt plus the input fuse plus the leads.

Both are the same idea in different currency: the instrument takes a share, and the size of that share is set by a ratio between the instrument and the circuit. The thermal, pressure and flow versions of this belong to a sibling article and are not re-derived here.

The divider that decides a voltage reading

Put a meter of input impedance Zm across a source whose internal or series impedance is Zs, and the meter reads:

Reading = true source voltage x Zm / (Zs + Zm)

That relationship is derived for a simple series path with one series impedance between the source and your probe points, which is the shape of nearly every field fault worth finding: a corroded splice, a pitted contact, an undersized run, a failing connection. It is the same divider whether the impedance is resistive or capacitive.

Read the consequence off the formula rather than memorising a rule. When Zm is enormously larger than Zs, the fraction is essentially 1 and the meter reads full source voltage no matter how bad the series impedance is. A 10 megohm meter and a 2 kilohm fault differ by a factor of five thousand, so the meter cannot see the fault at all. To make a series impedance visible in a voltage reading, the measuring impedance has to be within range of it, either because the real load is connected or because you deliberately lowered the meter's.

When loading is not worth thinking about

• The source impedance is tiny compared with the instrument. A 10 megohm meter on a branch circuit whose source impedance is a fraction of an ohm is not perturbing anything.
• The load stays connected while you read. The real load sets the operating point and the meter is a spectator alongside it. This is why readings taken under load answer questions that no-load readings cannot.
• You are comparing two readings taken the same way. A consistent loading error is common to both and largely drops out of the difference.

Where none of those hold, and especially where the load is disconnected or the node is high impedance by design, the loading is the measurement.

Low-impedance mode, and the two things it must never be

Many hand-held instruments have a low-impedance voltage mode that deliberately presents a few kilohms instead of tens of megohms. Its purpose is exactly the divider above: it drags the measuring impedance down into the same range as a fault or a coupling path so that path becomes visible.

Its main everyday use is ghost voltage. An unconnected conductor running alongside energized ones picks up voltage through the capacitance between them. That coupling is a very high impedance, so a 10 megohm meter reads substantial and entirely real voltage that can deliver almost no current. Low-impedance mode collapses it to near nothing, which is the answer you wanted.

Two hard limits on that mode:

• It is not a proving-dead instrument by itself. Proving a conductor dead is the live-dead-live sequence in NFPA 70E-2021, 120.5, under the electrical safety program that adopts it: verify the tester on a known live source, test the conductor, verify the tester again. A single reading of zero on any mode is not that sequence.
• It draws real current, and on some circuits that current does something. A few kilohms across a low-power control circuit, an electronic sensing input, a ground-fault detection circuit or a solid-state output can pull in a relay, operate a device, trip a protective function or damage the input. Know what the circuit drives before you load it, and do not use the mode on any circuit that serves a protective or combustion-safety function.

Burden voltage on the current side, and why the clamp exists

Every in-line current measurement subtracts burden voltage from the circuit. Whether that matters is again a ratio: burden voltage against the circuit's supply voltage. A few tenths of a volt in a 120 V or 240 V circuit is nothing. The same few tenths in a low-voltage control loop or a circuit with little voltage headroom can move the current you were trying to measure, and on a current loop it can push the transmitter out of its compliance range so the loop stops regulating altogether.

A clamp instrument avoids both problems at once. It does not break the circuit and it adds essentially no burden, because it reads the magnetic field around the conductor rather than passing the current through itself. It trades away resolution at low currents and it needs a single conductor in the jaw. That trade is why both instruments exist, and it is a selection question a sibling article owns.

The case: 24 volts at the wire, 2 volts at the coil

A contactor will not pull in. The control transformer is good. A tech pulls the coil wire off the terminal, puts the meter from that wire to common, and reads 24.0 V. Voltage present, coil must be bad, coil gets replaced. It does not fix it.

Work the divider. There is a corroded splice in the control run; call its resistance 2 kilohms. The meter's voltage input is 10 megohms.

With the coil wire off, the only path is source, splice, meter. The meter reads 24 V times 10,000,000 divided by 10,002,000, which is 23.995 V and displays as 24.0 V. The drop across the splice at that moment is 0.005 V. The fault is genuinely there and is genuinely invisible, because the meter is drawing a few microamps through it.

With the coil connected, the coil is the load. Call it 200 ohms. Now the same 24 V divides between 2 kilohms and 200 ohms, so the coil sees 24 times 200 divided by 2,200, which is 2.18 V, and the splice drops 21.8 V. The contactor sees roughly a tenth of its rated coil voltage and does nothing, which is exactly the complaint.

With the coil wire off and low-impedance mode selected, say the mode presents 3 kilohms. The reading becomes 24 times 3,000 divided by 5,000, which is 14.4 V, and the splice now drops 9.6 V. Nothing about the circuit changed. The instrument changed, and the fault became visible.

The two readings that would have found it directly. Measure across the suspect splice itself rather than at the coil: under load it reads 21.8 V, and with the load off it reads 0.005 V. That pair is the signature of a series resistance, and the difference between the two is the whole diagnosis. Or measure at the coil with the coil connected and energized, which is the general rule this case exists to teach: a voltage reading taken with the load disconnected tells you a path exists, not that it can carry anything.

What flips it. If the fault had been an open rather than a high resistance, the no-load reading would have been zero and the tech would have found it in one move. High-resistance faults are the ones this catches, and they are the majority of the connection faults a field tech meets, because corrosion and loosening produce resistance long before they produce an open.

The failure mode. A good coil is replaced, the fault stays, and the second visit starts from a worse position: the tech now has a "known good" part in place and a reading of 24 V on the record, both of which point away from the splice. The cost is not the part, it is that the evidence has been made harder to read.

How to verify you got this right

• Take the reading under load whenever the question is whether the circuit can do work. No-load voltage answers whether a path exists and nothing else.
• Measure across the suspect element, not around it. A drop across a connection under load is direct evidence; a voltage at a downstream terminal is inference.
• Compare a high-impedance and a low-impedance reading at the same point. If they differ substantially, you are looking at a high-impedance path, whether that is a fault or a coupling.
• Before selecting low-impedance mode, name what the circuit operates. If you cannot, do not load it.
• Check the function selector and the jacks agree before the tips move. The current input across a voltage source is the failure that ends with a hand injury rather than a wrong number.

References

• 29 CFR 1910.333(a)(1) and 1910.335(a) - the energized-work gate and the electrical protective equipment required for the exposure
• 29 CFR 1910.333(b)(2) and 29 CFR 1926.417 - electrical lockout in general industry and construction, required before breaking a circuit to insert an in-line ammeter
• NFPA 70E-2021, 120.5 - the live-dead-live proving sequence, binding through the employer's electrical safety program or an adopting jurisdiction
• See related: What a Meter Category Rating Is Protecting You From; The Measurement That Changes What It Measures; How to Choose the Right Instrument for the Question