Current Transformer vs Rogowski Coil: AC Current Sensing for Power Electronics
Last updated 7 July 2026 · 6 min read
Direct Answer
A current transformer (CT) uses a laminated iron or ferrite core wound with a secondary coil around the current-carrying conductor, transforming a changing primary current into a proportionally scaled secondary current — simple, self-powered (no external excitation needed), and low-cost, but it saturates at high current and cannot measure DC, since it depends entirely on a changing magnetic flux to induce a secondary voltage. A Rogowski coil is an air-core coil (no iron core to saturate) wrapped around the conductor whose output voltage is proportional to the rate of change of current (di/dt), requiring an external integrator circuit to recover a current-proportional signal — this trades away the CT's self-powered simplicity for a much wider dynamic range, no saturation at high current, and a flexible, often clip-on form factor. Both are AC-only, galvanically isolated current sensing techniques used in energy metering, motor drive protection, and power electronics test and measurement, distinct from the DC-capable Hall-effect and shunt-resistor techniques covered elsewhere on this site.
Detailed Explanation
DC-capable current sensing techniques — shunt resistors and Hall-effect ICs — are covered in detail in Hall-effect current sensor vs shunt resistor, which names current transformers only once, in passing, to contrast their AC-only limitation against Hall-effect sensing's DC capability. This page covers the two dominant AC-only, non-contact current sensing techniques directly: the current transformer and the Rogowski coil.
Current Transformer (CT)
A current transformer works exactly like a power transformer, with the current-carrying conductor itself acting as a one-turn primary winding passing through the CT's core, and a secondary winding (often hundreds or thousands of turns) wound around a laminated iron or ferrite core. The changing primary current induces a proportionally scaled current in the secondary winding, following the transformer's turns ratio — a 1:1000 CT with 5 A flowing in the primary conductor produces 5 mA in the secondary.
Key characteristics:
- Self-powered. No external excitation supply is needed — the CT derives its output entirely from the changing primary current.
- Requires a burden resistor. The secondary current is converted to a measurable voltage across a burden resistor sized specifically for the CT — see the FAQ above for why this sizing matters.
- Saturates at high current or low frequency. Exceeding the core's rated volt-time product (from excessive current, too high a burden resistance, or operation below the core's designed frequency range) drives the core into saturation, where the output becomes nonlinear and no longer accurately represents the primary current.
- Cannot measure DC. See the FAQ above — a CT depends entirely on Faraday's law induction from a changing flux.
Rogowski Coil
A Rogowski coil is a helical coil wound on a non-magnetic (air or non-ferrous) core, typically formed into a flexible loop that clips around the conductor without needing to break the circuit to install it — a genuine practical advantage over a solid-core CT in retrofit or test-and-measurement situations.
Key characteristics:
- No iron core, so no saturation. Because there's no ferromagnetic core to saturate, a Rogowski coil has an extremely wide dynamic range, from small currents up to many kiloamps, without the nonlinearity a CT exhibits near saturation.
- Output is proportional to di/dt, not current. This is the coil's defining trade-off — see the FAQ above for why an integrator circuit is required to recover a current-proportional signal, and what happens if it's skipped.
- Needs external power for the integrator. Unlike a self-powered CT, a Rogowski coil measurement system needs a powered integrator circuit (built into most commercial Rogowski current probes and transducers).
- Flexible, clip-on form factor. Many commercial Rogowski coils are designed as a flexible loop that wraps around a conductor or busbar without disconnecting it — valuable for test and measurement applications and for large or irregularly shaped conductors that don't fit a fixed CT window.
Choosing Between Them
| Factor | Current transformer | Rogowski coil |
|---|---|---|
| Self-powered | Yes | No — needs an integrator supply |
| Saturation at high current | Yes, above core rating | No — no iron core to saturate |
| Dynamic range | Limited by core saturation | Very wide |
| Installation | Usually fixed-window (conductor passes through) | Often flexible, clip-on |
| Output signal | Current-proportional directly | di/dt — needs integration |
| Typical use | Energy metering, protection relays, fixed installation monitoring | Power quality analysers, motor drive test, high-current or irregular-conductor measurement |
| DC measurement | No | No |
Fixed, cost-sensitive, high-volume applications (energy meters, protection relays, motor drive overcurrent protection) generally favour current transformers for their simplicity and self-powered operation. Test and measurement equipment, power quality analysis, and applications needing very wide dynamic range or a flexible clip-on form factor favour Rogowski coils.
Design Considerations
- Verify the CT's rated burden before finalising the secondary-side circuit. Adding a downstream amplifier, ADC input network, or protection circuitry all add impedance in series with the CT's secondary — the total burden the CT sees must stay within its rated limit, not just the resistor you intentionally add. See the FAQ above for the saturation consequence of exceeding it.
- Check the CT's rated frequency range against the actual signal. A CT sized for mains-frequency (50/60 Hz) energy metering may not accurately measure higher-frequency content — switching-frequency ripple current in a power electronics application needs a CT (or Rogowski coil) explicitly rated for that bandwidth.
- Confirm the integrator's own bandwidth and drift for a Rogowski coil design. Since the current-proportional signal only exists after integration, the integrator circuit's own frequency response and DC offset drift directly become part of the overall measurement's accuracy — this is not a passive, "set and forget" stage the way a CT's burden resistor is.
Zeus Design's electronics engineering team designs current sensing circuits for motor drive protection, energy metering, and power electronics test applications as part of complete product electronics design.
Common Mistakes
- Assuming a CT or Rogowski coil can measure DC or low-frequency current. See the FAQ above — both techniques require a changing flux and produce no usable output for steady DC; use a Hall-effect sensor or shunt resistor instead for DC or low-frequency measurement (see Hall-effect current sensor vs shunt resistor).
- Operating a CT open-circuited on the secondary (no burden connected). With no burden resistor to limit it, an open-circuited CT secondary can develop a dangerously high voltage across the winding, since the transformer tries to maintain its ampere-turn balance with no path to do so safely — always keep a CT's secondary loaded, even temporarily during test or calibration.
- Skipping the integrator on a Rogowski coil and treating the raw output as a current reading. Covered in the FAQ above — the raw di/dt signal is not current, and using it as if it were produces a plausible-looking but incorrect waveform, particularly for non-sinusoidal currents.
- Selecting a CT or Rogowski coil rated for mains frequency in a switching power electronics application. The core material and coil design suited to accurate 50/60 Hz measurement may not have adequate bandwidth or accuracy for switching-frequency current content — check the rated frequency range against the actual application, not just the nominal current rating.
Frequently Asked Questions
- Why does a current transformer need a burden resistor?
- A current transformer is designed to operate into a very low secondary-side impedance — ideally close to a short circuit, from the transformer's perspective. The burden resistor is the load placed across the CT's secondary winding to convert the secondary current into a measurable voltage. Sizing it matters in both directions: too high a burden resistance forces the CT core toward saturation and distorts the output (because the secondary voltage needed to drive current through a larger burden approaches the core's voltage-time saturation limit), while too low a burden gives a very small, harder-to-measure voltage signal. The CT manufacturer's datasheet specifies a rated burden (or a maximum burden resistance) for accurate operation — exceeding it is one of the most common causes of an unexpectedly nonlinear or clipped CT reading.
- Why does a Rogowski coil need an external integrator, and what happens if you skip it?
- A Rogowski coil's raw output voltage is proportional to di/dt (the rate of change of current), not to the current itself, because the coil has an air core rather than a saturable iron core with a defined turns ratio like a CT. Without integration, the raw signal is only useful for detecting fast edges (some protection relays and fault detectors use the raw di/dt signal directly for exactly this reason) but does not represent the actual current waveform. An analog or digital integrator circuit — typically an op-amp integrator, though modern coil-and-electronics packages often integrate this internally — recovers a signal proportional to current itself from the di/dt output. Skipping the integrator on a design that assumes a current-proportional signal produces a reading that looks plausible for a sinusoidal AC waveform (a 90-degree phase-shifted, differently-scaled version of the true current) but is not actually a valid current measurement, and is especially misleading for non-sinusoidal or fast-transient waveforms.
- Why can neither a current transformer nor a Rogowski coil measure DC current?
- Both techniques depend entirely on a changing magnetic flux to produce their output signal — a CT induces a secondary voltage via Faraday's law of electromagnetic induction, which requires a time-varying flux; a Rogowski coil's output is explicitly proportional to di/dt, the rate of change of current. A steady DC current produces a constant magnetic field with no time variation, so neither technique produces any usable output at all for pure DC. This is the same fundamental limitation named in the [Hall-effect current sensor comparison](/questions/hall-effect-current-sensor-vs-shunt-resistor) — Hall-effect and shunt-resistor sensing are the appropriate choices whenever DC or low-frequency current measurement is required; CTs and Rogowski coils are reserved for genuinely AC applications (mains-frequency energy metering, motor drive AC-side monitoring, switching-frequency ripple measurement).
References
Related Questions
Hall-Effect Current Sensor vs Shunt Resistor: Which Should You Use?
Hall-effect current sensors give galvanic isolation but worse drift and bandwidth than shunt resistors. Learn when the isolation trade-off is worth it.
How Do You Measure Current with a Shunt Resistor?
Covers shunt resistor selection, low-side vs high-side sensing, INA219/INA226 I2C monitors, Kelvin connection layout, and calibration register calculation.
How Do You Use the INA219 and INA226 I2C Current-Sense ICs?
Wire, calibrate, and configure the INA219 and INA226 I2C current monitors — register map, calibration formula, ALERT pin setup, and a comparison of both ICs.
Sensor Signal Conditioning Basics
Covers signal conditioning for sensors: op-amp gain, offset, 4-20mA interface, anti-aliasing filter design, ADC input protection, and two-point calibration.
How Do You Design a Constant-Current LED Driver Circuit?
LEDs need constant current, not constant voltage. Covers series resistors, linear and switching driver ICs, PWM dimming, and multi-string current hogging.
How Do You Interface a MEMS Barometric Pressure Sensor?
MEMS barometric pressure sensor interfacing: I2C/SPI digital sensors, factory calibration coefficients, temperature compensation, and altitude calculation.
Related Forum Discussions
ICM-42688-P INT1 never fires after FIFO setup — FIFO count stuck at 0x0000 but WHO_AM_I reads back correctly
Working on a motion data logger — ICM-42688-P on an STM32F4 via SPI4 at 4 MHz. SPI seems fine: WHO_AM_I at register 0x75 reliably returns 0x
LDO overheating and shutting down on a 2S Li-ion powered board
Got a board running off a 2S Li-ion pack (7.4V nominal, up to 8.4V fresh off the charger) feeding an LDO down to 3.3V for the MCU and a coup