As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A Hall effect clamp puts a semiconductor element into a gap in an iron core, where it responds to the magnetic field itself. That means it reads direct current, alternating current and anything in between, at the price of saturating at the top of its range, drifting with temperature and needing to be zeroed. A Rogowski coil responds to the rate at which the field changes instead, so it cannot see a settled direct current at all, but nothing in it saturates, it stays linear from a handful of amps to several thousand, it weighs almost nothing, and it threads around a conductor that no jaw would ever close over. Choose Hall when there is a direct component or when the conductor is small and reachable. Choose Rogowski when the current alternates and the conductor is large, crowded or carrying fast switching edges.
Two ways to turn a magnetic field into a number
A Hall element is a thin slab of semiconductor with a known current pushed along its length. Put a magnetic field across the face of the slab and the moving charge carriers are deflected sideways, piling up on one edge until the electric field they create balances the deflection. The voltage between the two side contacts is then proportional to the flux density. It is a static effect: hold the field constant and the voltage stays put indefinitely. Wrapping the element in an iron core with a narrow gap concentrates the field from the conductor into the gap, multiplying the signal and shielding it from fields that are not the one you want.
The other method throws the iron away. A Rogowski winding encircles the conductor with air inside the turns, and its output is proportional to how fast the current is changing rather than to how large it is, with an integrator stage afterwards converting that back into something that reads in amps. Because nothing in the loop can saturate, the span between the smallest and the largest current it will handle linearly is enormous. The ETCR6920B flexible coil clamp multimeter quotes alternating current from 0 mA up to 10,000 A on a coil 300 mm around, a span no gapped core of a sane size could cover, and the same listing also gives alternating and direct voltage up to 750 V and 1000 V respectively on the probe jacks, resistance to 2 kilohms and continuity.
Saturation, offset and the price each method pays
Iron has a ceiling. Push enough current through the conductor and the flux density in the core reaches the point where the material will not magnetise any further, somewhere in the region of one and a half to two tesla for the silicon steels used in instrument cores. Beyond that the relationship between current and gap field flattens out and the reading quietly stops being linear. Core size, gap width and the maximum range on the dial are all tied together by that ceiling, which is why a clamp rated for very large currents is physically large.
Iron also remembers. After a core has carried a heavy current it retains some magnetisation, and the Hall element faithfully reports that residue as a current that is not flowing. Add the element offset, which is a property of the semiconductor and moves with temperature, and you have the two reasons a direct current clamp carries a zero control. Neither of those failure modes exists in a coil with no core in it. What a coil has instead is a small output that must be amplified, an integrator whose own drift has to be managed, and a dependence on geometry: the loop assumes the conductor sits roughly in the middle and that the winding is uniform, and the clasp where the loop closes is the one place where it is not.
Bandwidth, inrush and the shape of the waveform
Frequency response separates the two methods as sharply as direct current does. A gapped core starts to lose energy to eddy currents and hysteresis as the frequency climbs, so the usable band of a typical Hall clamp is measured in a few kilohertz. A coil with air inside it has none of those losses and keeps responding far beyond that, which is why these probes turn up on drive outputs, welding circuits, capacitor discharge rigs and anywhere else the interesting part of the waveform is the edge rather than the average.
| What you are measuring | Hall effect in a gapped core | Rogowski coil with an integrator |
|---|---|---|
| Settled current in a battery or solar string | Reads it directly, after zeroing | Reads zero, by construction |
| Motor inrush on an alternating supply | Captures it if the peak is inside the core range | Captures it with no saturation risk |
| Several thousand amps on a busbar | Needs a very large core, or is simply out of range | Comfortable, since there is nothing to saturate |
| Fast switching edges from a drive output | Rolls off as core losses rise | Strong, because output grows with rate of change |
| Milliamp leakage in a large cable | Purpose built jaws resolve to microamp level | Limited by amplifier noise at the low end |
| Rectified supply with ripple on top | Shows the whole waveform including the average | Shows the ripple and discards the average |
When the current reading is not the measurement you need
Both methods answer the same question, which is how much current is flowing. Neither answers the question that usually matters on a fault call, which is where the energy is being wasted. A loose lug carries exactly the same current as a tight one. The difference is the resistance of the joint and therefore the heat it produces, and heat is a temperature measurement rather than a current measurement.
This is where a thermal imager earns its keep alongside a clamp. The FLIR One for Android states that it upscales the image to 240 by 180 from an 80 by 60 sensor and blends thermal with visible detail in real time, while the FLIR One Pro states 480 by 360 upscaled from 160 by 120. Both listings quote temperature accuracy within 3 °C or 5 percent with the unit between 15 °C and 35 °C and the scene between 5 °C and 120 °C. Neither states a noise equivalent temperature difference, the figure that governs how small a temperature step the sensor can distinguish, so that is not stated in the listing bullets and has to come from the data sheet. Those differences are picked apart in the FLIR One review and the FLIR One Pro review, and the resolution step between a phone attachment and a standalone instrument is the subject of the FLIR One Pro against FLIR Edge Pro comparison.
The practical workflow pairs them. Clamp first to establish that a circuit is loaded as expected, then scan for the joint that is hotter than its neighbours at that load. A hot joint at low load is more serious than a warm one at full load, and you cannot judge that without both numbers.
Matching the sensor to the conductor and the signal
Five questions settle the choice, in this order. Does the current have a direct component, which immediately rules out a coil. What is the largest current you need to read, which sizes a core or rules one out. What is the smallest current you need to resolve, which is where a coil struggles and a dedicated jaw shines. What does the conductor physically look like, since a flexible loop goes where a jaw cannot. And how fast are the edges in the waveform, because that is where the coil pulls ahead decisively.
Answer those and the instrument almost chooses itself, with one caveat worth repeating: a current figure alone is a thin diagnosis. Pair it with a temperature picture of the same connection and the two together tell a story that neither tells alone, which is the argument running through the comparison between the FLIR One Pro and the Topdon TC002C. Buy the sensing element for the electrical question you are actually asking, then make sure something in the bag answers the thermal one.
Frequently Asked Questions
Which method is more accurate?
Neither wins outright, because they fail in different places. A Hall effect clamp is typically specified as a percentage of reading plus a number of counts, and its weak point is the low end, where offset and remanence matter most relative to the reading. A Rogowski probe keeps the same percentage accuracy across a far wider span because nothing saturates, and its weak point is positioning, since where the conductor sits inside the loop changes the coupling slightly.
Why does a Hall effect clamp need a zero button?
Two reasons add together. The semiconductor element produces a small output even with no field present, and that offset moves with temperature. Separately, the iron core keeps a little residual magnetism after it has carried a large current, which the element reads as a field that is not there. Pressing zero with the jaws closed and empty cancels both at once, and the habit matters most when the current you are about to measure is small.
Can a Rogowski probe measure a very small alternating current?
It can, but the low end is where it works hardest. The coil output is small to begin with, so the amplifier noise floor sets a practical limit, and stray fields from neighbouring conductors become significant when the signal is weak. For milliamp level leakage work, a purpose built rigid jaw with a high permeability core and a tightly closing joint usually resolves lower, which is why leakage instruments are built that way.
Does a Rogowski coil work on high frequency drive output?
That is one of its strongest cases. There is no core to introduce losses or saturation at high frequency, and the coil output naturally rises with frequency, so fast switching edges arrive with plenty of signal. The limit is set by the integrator and the cable rather than the winding. A gapped iron core, by contrast, starts losing energy to the core itself as frequency climbs.
Is a flexible probe safer on a live conductor?
It can be easier to apply, which is a different thing from safe. A flexible loop threads behind and between conductors where a rigid jaw would have to be forced in, so there is less chance of bridging two phases with a tool. The insulation rating, the measurement category and the state of the conductor still decide whether the work should be happening live at all.