A current transformer (CT) is an instrument transformer designed to produce a secondary alternating current strictly proportional to the primary current it measures, providing galvanic isolation between high-voltage mains and low-voltage metering. In a real circuit or installation, a CT changes the game by allowing you to safely measure, log, or trigger protective relays on a 400A busbar using 5A or milliamp-level signals without ever interrupting the main power flow or exposing your microcontroller to lethal mains voltage.
People commonly confuse CTs with standard step-down power transformers or Hall-effect sensors. A power transformer is a constant-voltage device meant to transfer real power (watts) to a load. A CT, conversely, acts as a constant-current source meant to transfer a measurement signal. Furthermore, while Hall-effect sensors (like the ACS712) can measure both AC and DC using semiconductor physics, CTs only work with AC and rely entirely on magnetic induction.
The Core Principle: Ratios, Burden, and Magnetic Flux
At its core, a CT operates on the principle of ampere-turn balance. The primary winding (often just a single heavy cable passing through the center window) creates a magnetic flux in the iron or ferrite core. The secondary winding, which has many turns of fine wire, generates a current that creates an opposing flux. Under normal operation, the net flux in the core remains very low because the secondary ampere-turns almost perfectly cancel the primary ampere-turns.
The relationship is defined by the turns ratio. If a CT has a ratio of 400:5, it means 400 amps on the primary yields 5 amps on the secondary. However, a CT cannot drive an infinite impedance. It requires a burden resistor to convert that secondary current into a measurable voltage. The burden must be sized carefully: too high, and the core saturates (introducing non-linearity and dangerous voltages); too low, and your signal-to-noise ratio drops into the mud.
Worked Numeric Example: Sizing a Burden Resistor for an ESP32
Let us build a smart energy monitor using an ESP32 DevKit v1 and a popular YHDC SCT-013-000 split-core CT. The SCT-013-000 has a ratio of 100A:50mA (a 2000:1 turns ratio) and no internal burden resistor. We need to interface this with the ESP32's 12-bit ADC.
- Identify ADC Limits: The ESP32 ADC maxes out at 3.3V, but it becomes highly non-linear above 3.1V. Because AC current swings positive and negative, we must bias the signal to a 1.65V midpoint using a voltage divider. This leaves a maximum peak voltage swing of roughly 1.55V (1.65V to 3.2V) to stay in the linear region.
- Calculate Peak Secondary Current: At the 100A primary maximum, the RMS secondary current is 50mA (0.05A). The peak current is
0.05A × √2 (1.414) = 0.0707A. - Calculate Burden Resistance: Using Ohm's Law (R = V / I), we divide our max peak voltage by our peak current:
1.55V / 0.0707A = 21.92 Ω. - Select the Component: We choose the nearest standard resistor value, which is 22 Ω. For safety and to handle the power dissipation (
I²R = 0.05² × 22 = 55mW), a standard 1/4W 1% metal film resistor is more than adequate.
Bench Pro-Tip: The ESP32's internal ADC is notoriously noisy and non-linear. For revenue-grade or highly accurate solar monitoring, bypass the internal ADC and route the burden voltage into an external ADS1115 16-bit I2C ADC, which provides stable, predictable sampling via its programmable gain amplifier.
Where You Meet CTs in Practice
You interact with CT-derived data constantly, even if you do not see the physical components. In modern electrical infrastructure, they are the primary eyes and ears of protection and monitoring systems.
| Application | Typical CT Form Factor | Secondary Output | Real-World Example |
|---|---|---|---|
| Home Energy Monitors | Split-core, snap-on | 50mA or 333mV | Emporia Vue, Sense Home Energy |
| Solar Inverter Metering | Solid-core, DIN-rail mount | 5A or 1A | Fronius Smart Meter, SolarEdge |
| Industrial Motor Protection | Window-type, epoxy potted | 5A (or 1A for long runs) | SEL-710 Motor Protection Relay |
| Utility Revenue Metering | Bar-type or bushing mount | 5A (0.2 Accuracy Class) | Landis+Gyr grid edge meters |
Scenario Walkthrough: The Open-Circuit Catastrophe
To understand why CT safety rules are written in blood, consider this real-world jobsite scenario.
- The Setup: An industrial facility is monitoring a 200A, 3-phase HVAC compressor motor. The panel uses a standard 400:5 window CT with the secondary wired to an analog panel ammeter via 14 AWG wire.
- The Numbers: The compressor is running under heavy load, drawing 180A on the primary. According to the ratio, the secondary current is
180A × (5 / 400) = 2.25A. This 2.25A flows safely through the ammeter's low-impedance shunt. - The Action: A maintenance technician notices the analog meter is failing and decides to swap it for a digital display. Assuming it is just a low-voltage signal wire, they pull the secondary wire off the terminal block while the motor is still running.
- What Went Wrong: By removing the wire, the technician created an open circuit on the secondary. With no secondary current to oppose the primary magnetic field, the core instantly saturates. According to Faraday's Law of Induction (
V = -N × dΦ/dt), the massive, unopposed rate of change of magnetic flux induces a colossal voltage spike across the open secondary terminals—often reaching 2,000V to 5,000V.
The 14 AWG wire insulation instantly arcs over and vaporizes. The CT core overheats and shatters from the magnetic forces, sending shrapnel into the panel. The technician, holding the bare wire, faces a lethal shock hazard, and the resulting arc-flash triggers the main upstream breaker, blacking out the facility.
As detailed in All About Circuits' guide on instrument transformers, a CT secondary must always be treated as a live, high-energy source when primary current is present. If you must remove a meter, you must first install a shorting block or jumper wire across the secondary terminals.
FAQ: Current Transformers in the Field
Can I use a CT to measure DC current from a battery bank?
No. CTs rely on a changing magnetic field (AC) to induce a secondary current. A steady DC current will simply magnetize the core once and produce zero secondary output. For DC battery banks (like a 48V LiFePO4 server rack), you must use a shunt resistor with an isolated amplifier or a Hall-effect sensor like theLEM DHAB s/14.
What does 'Accuracy Class' mean on a CT datasheet?
Accuracy class defines the maximum percentage error at the rated current. A 'Metering Class 0.3' CT will be accurate to within 0.3% from 10% to 100% of its rated current, making it ideal for utility billing. A 'Protection Class 5P' CT is less precise at normal loads but is designed to remain linear (not saturate) during massive fault currents (e.g., 20x rated current) so the protective relay can trip accurately. For a deep dive into saturation curves and relay coordination, refer to Electronics Tutorials' comprehensive breakdown.
Does the direction of the wire through the CT window matter?
Absolutely. CTs have polarity marks (usually a white dot, an arrow, or 'P1/P2' and 'S1/S2' labels). If you are measuring single-phase power, reversing the CT will invert the current waveform relative to the voltage, causing your power factor and real power (Watts) calculations to read negative. In 3-phase systems, mixing up the polarity on one leg will cause vector summation errors, leading to wildly inaccurate total power readings.






