What a Clamp Measurement Can and Cannot Tell You

Why this matters

A clamp reading is the fastest number in the trade and the one most often over-read. It arrives in two seconds, it has three or four digits on it, and a tech will condemn a motor, clear a circuit, or sign off a service call on the strength of it. The digits are not the problem. The problem is that the number answers exactly one narrow question, and the tech who does not know which question was asked will answer a different one with it.

This card is the negative space. It is a list of what is genuinely not in that number, and why each absence is invisible on the display rather than flagged as an error. The related how-to on reading a clamp under load owns the procedure. What this card owns is the boundary of the claim you are allowed to make afterward.

Before the jaw goes anywhere near a conductor

Clamping around an insulated conductor inside an enclosure means the cover is off and energized parts are exposed. Under 29 CFR 1910.333(a)(1) you de-energize before working on or near exposed energized parts unless de-energizing introduces additional or increased hazards or is infeasible because of equipment design or operational limitations, and a current measurement that only exists under load is one of the recognized cases where the work is done live. That is a determination you make and can defend, not an assumption you make because the reading is easier that way.

If you go live: work under your employer's electrical safety program, use the shock and arc-flash boundaries and the PPE that program specifies (NFPA 70E-2021 is the usual basis, and it binds you through your employer's program or through your contract, in whatever edition that program has adopted, never on its own authority), and provide and use electrical protective equipment as required by 29 CFR 1910.335(a).

Check the measurement category and voltage rating marked on the meter and on the jaw itself, not just on the leads. That marking is established under IEC 61010-1 as applied by the listing body whose mark is on your instrument, and it binds through the listing, which is what your inspector and your insurer will look at. A clamp rated for a lower category than the point you are standing at is the wrong tool regardless of whether it will physically fit. Inspect the jaw faces for cracks, contamination and a clean close before every use: a jaw that does not close fully reads low and does not tell you it is doing so.

What the jaw actually senses

A clamp does not measure the current in a wire. It measures the net magnetic effect of everything enclosed by the jaw. That is a single sentence, and almost every limitation below falls out of it.

On a healthy two-wire branch circuit, the supply conductor and the return conductor carry the same current in opposite directions. Enclose both and the net is essentially zero. Enclose one and you get that conductor's current. The instrument has no way to know which of those two situations it is in. It reports the net either way, and a near-zero reading looks identical whether the circuit is off, the circuit is balanced and healthy with both conductors in the jaw, or the jaw did not close.

That property is also the clamp's single best trick, and it is worth stating plainly because it is under-used: enclosing all the current-carrying conductors of a circuit at once and reading anything other than a near-zero net is direct evidence that current is returning by some path outside the jaw. That is a leakage or a parallel-path finding you cannot get any other way with a hand tool.

What the number does not contain

Six absences, and none of them show on the display.

Circuit identity. The reading tells you a magnitude, not whose current it is. In a bundle, in a wireway, on a shared neutral, or where a conductor loops back through the same enclosure, you can get a perfectly stable number off the wrong conductor. Trace the conductor to both ends before you attach a meaning to its current.

Direction and phase relationship. A basic clamp reports magnitude only. It cannot tell you whether two currents are in phase, whether one leg is leading, or which way power is flowing at that instant. Two legs reading the same magnitude are not evidence of a balanced load, only of equal magnitudes.

Power. Current is not power. A motor drawing current at poor power factor moves more amps for the same real work, and the clamp reports the amps. If the question is consumption, or whether the load is doing the work it should, current alone answers it wrongly in the direction that looks worse than reality: a motor at poor power factor moves more amps for the same real work, so route that question to a measurement that captures voltage and phase angle together. Heating in the supply conductors is the exception and runs the other way. That goes with the square of the current regardless of phase angle, so the clamp reading is exactly the right number for it, and a high current on a poor-power-factor load is heating the conductors at its full value whatever the real power turns out to be. Route that question to a measurement that captures voltage and phase angle together.

Waveform content, unless the meter is true-RMS. An averaging meter is calibrated to read correctly on a clean sine wave and is wrong on every other shape, in a direction that depends on the shape: it reads low on the peaky current a rectifier front end draws, and high, by up to 11 percent on a square wave, on a flat-topped one. It does not warn you either way. If the load has power electronics in it, use a true-RMS instrument and confirm the specification sheet says so. The general rule is that if the load has power electronics in it, use a true-RMS instrument and confirm the specification sheet says so rather than assuming the price of the tool implies it.

DC, on an AC-only clamp. A current-transformer clamp responds to a changing field and is blind to steady DC. Reading zero on a DC circuit with the wrong clamp is a false negative that looks exactly like a de-energized circuit. Only a Hall-effect type reads DC, and it needs its zero set before each reading because that zero drifts.

Anything below the meter's low-end floor. This is the one that costs the most, and it gets its own section.

The resolution floor is not the number of digits

Three properties get conflated constantly, and buying or trusting the wrong one is the direct consequence. Resolution is the smallest change the display can show. Accuracy is how far the reading may sit from the true value. Repeatability is how much the reading moves when nothing about the world has changed. A clamp can have excellent resolution, poor accuracy, and good repeatability all at once, and a tech reading four digits will assume all three are good. The related card on calibration scheduling owns the distinction in general terms; what matters here is the arithmetic it produces on a clamp.

A clamp's accuracy specification is almost always written as a percentage of reading plus a fixed number of counts. The percentage term is proportional and stays modest. The counts term is fixed by the range you are on, so it becomes a larger and larger share of the reading as the reading gets smaller. A large-range clamp measuring a small current is the classic case of an instrument that displays a number it cannot support.

Accuracy classes vary widely between instruments and between ranges on the same instrument. Read the specification sheet for the range you are actually on rather than assuming one figure covers the tool.

Worked example: the same motor, the same clamp, two ranges

A small blower motor with an illustrative nameplate full-load current of 5.0 A. The question is whether it is running near nameplate or well under it, because that changes whether you keep investigating the motor or move to the driven load.

Say the clamp's specification sheet lists, for its 600 A range, an accuracy of plus or minus 2 percent of reading plus 5 counts, with 0.1 A resolution on that range. On its 60 A range it lists the same 2 percent of reading plus 5 counts, with 0.01 A resolution. Those figures are illustrative and stand in for the ones on your own instrument's sheet.

You clamp on the 600 A range because that is where the switch was left, and you read 4.2 A.

• The proportional term: 2 percent of 4.2 A is 0.084 A.
• The counts term: 5 counts at 0.1 A per count is 0.5 A.
• Total uncertainty: 0.084 plus 0.5 is 0.584 A, which against a 4.2 A reading is about 14 percent.
• The reading therefore supports a true value anywhere from about 3.6 A to about 4.8 A.

That band straddles the decision. At 3.6 A the motor is running well under nameplate and your attention belongs on the driven load. At 4.8 A it is within a hair of nameplate and your attention belongs on the motor and its supply. The display showed you 4.2 and it was not lying, but it could not resolve the question you brought to it.

Now switch to the 60 A range and re-read the same motor at the same load.

• The proportional term is unchanged: 2 percent of 4.2 A is 0.084 A.
• The counts term: 5 counts at 0.01 A per count is 0.05 A.
• Total uncertainty: 0.084 plus 0.05 is 0.134 A, about 3.2 percent of the reading.
• The reading now supports a true value from about 4.07 A to about 4.33 A.

Same jaw, same motor, same instant, and the second reading answers the question while the first does not. Nothing about the display told you which one you were holding. The number of digits was identical.

The failure mode. The tech who took the first reading writes "4.2 A, under nameplate, motor fine" and moves on. Six weeks later the motor is warm and the customer is calling again. The record shows a number, so nobody re-measures, and the wrong branch of the investigation stays closed on the strength of a reading that never had the resolving power to close it.

What would change this. If the question had been "is this circuit carrying anything at all" rather than "is this motor near nameplate", the 600 A range reading answers it completely and the range choice does not matter. The band you need is set by the decision, not by the instrument. Decide what difference would change your action, then pick the range that can resolve it.

What this example does not settle. Even the 4.2 A on the correct range still tells you nothing about power factor, so it cannot tell you that the motor is doing 4.2 A worth of useful work. It also assumes one conductor in the jaw and true-RMS response. Both of those are assumptions you made, not measurements you took.

Making the reading portable to next week

A clamp reading is evidence about one instant under one load state. Write down what makes it re-checkable: the conductor you clamped and how you identified it, the range the meter was on, whether the meter is true-RMS, what the load was doing at the time (starting, steady, staged up, cycling), and the supply voltage if you have it. Without those, the number cannot be compared to a later one, because you will not know whether a change in the number is a change in the equipment or a change in how you took it.

How to verify you got this right

Before you act on a clamp reading, answer these four out loud:

1. What is in the jaw? Name the conductor and how you know. If the answer is "the one that was easiest to reach", trace it first.
2. Is the counts term small against the reading? Compute it on the range you are on. If the uncertainty band spans your decision, change range and re-read.
3. Does the reading move when the jaw moves? Reposition the JAW so the conductor sits nearer its centre, and re-read. Never handle an energized conductor to improve a reading: moving live conductors is not the measurement the energized-work determination covered, and if the conductor genuinely has to move, that is de-energized work under 29 CFR 1910.333(b)(2). A reading that shifts noticeably with position is telling you about coupling or an adjacent conductor, not about the load.
4. Would a different question need a different instrument? If you are about to say anything about power, heating, or work done, current alone did not establish it.

References

• 29 CFR 1910.333(a)(1) - general requirement to de-energize before working on or near exposed energized parts, with the infeasibility and increased-hazard exceptions
• 29 CFR 1910.335(a) - use of electrical protective equipment for work on or near energized parts
• IEC 61010-1 measurement categories, as applied through the listing mark on your own instrument and leads
• Manufacturer specification sheet for your clamp: accuracy by range, true-RMS response, and DC capability
• See related: Reading a Clamp Meter Under Load; Clamp Meter vs Multimeter: Which Tool to Reach For; The Calibration Schedule Worth Keeping