A current transformer (CT) is a type of instrument transformer designed to step down high primary AC currents to a safe, standardized secondary current (typically 5A or 1A) for metering and protection. In a real circuit or installation, it changes the measurement paradigm entirely, allowing standard 5A panel meters, PLC analog inputs, or ESP32 ADC pins to safely monitor 100A to 5000A+ mains circuits without any galvanic connection to lethal line voltages. Makers and junior technicians commonly confuse CTs with Hall-effect sensors (which measure both AC and DC) or Voltage Transformers (PTs), a mistake that often leads to catastrophic open-circuit failures because a CT behaves as a constant current source, not a voltage source.
Core Materials and Winding Topologies
The physical construction of a current transformer revolves around a toroidal magnetic core wrapped with hundreds or thousands of turns of fine enameled copper wire (the secondary winding). The primary "winding" is often just the single straight busbar or heavy gauge wire passing through the center of the torus. Because the primary current is dictated entirely by the external load (the grid or a motor), the CT core must be engineered to handle the resulting magnetic flux without saturating prematurely.
The choice of core material dictates the CT's accuracy, cost, and physical size. Below is a specification breakdown of the four most common core materials used in modern CT construction, ranging from cheap hobbyist sensors to utility-grade metering.
| Core Material | Initial Permeability ($\mu_i$) | Saturation Flux Density ($B_{sat}$) | Typical Accuracy Class | Primary Use Case & Cost |
|---|---|---|---|---|
| Grain-Oriented Silicon Steel | 1,500 - 3,000 | ~2.0 Tesla | Class 1.0 to 5P (Protection) | Industrial switchgear, high fault-current tolerance. Low cost. |
| Nickel-Iron (Permalloy) | 10,000 - 20,000 | ~0.8 Tesla | Class 0.5 to 0.2 | Revenue metering, precision PLC feedback. Medium cost. |
| Nanocrystalline | 30,000 - 100,000 | ~1.2 Tesla | Class 0.2S to 0.1 | Utility smart meters, wide dynamic range. High cost. |
| Manganese-Zinc Ferrite | 2,000 - 10,000 | ~0.4 Tesla | Unclassified (Hobbyist) | DIY split-core sensors (e.g., SCT-013), high-frequency SMPS. Very low cost. |
Notice the trade-off: Silicon steel can handle massive magnetic flux (high $B_{sat}$) before saturating, making it ideal for protection relays that must survive 10,000A short-circuit faults. However, its low permeability makes it inaccurate at very low currents. Nanocrystalline cores offer extreme permeability for pinpoint accuracy at 1A loads, but will saturate and clip the waveform during a massive fault event unless specifically gapped.
The Math: Turns Ratio, Burden, and a Worked Numeric Example
The fundamental operating principle relies on Ampere-Turns balance: $I_p \times N_p = I_s \times N_s$. If you have a single primary busbar ($N_p = 1$) and you want to step 400A down to a standard 5A secondary, your turns ratio must be 80:1. Therefore, the secondary winding requires exactly 80 turns ($N_s = 80$).
However, a CT cannot operate without a burden—a resistive load connected across the secondary terminals to convert the induced current into a measurable voltage. Think of a CT like a positive-displacement water pump forcing a fixed flow rate (current) through a pipe. The burden resistor is a valve restricting that flow. If you close the valve entirely (open circuit), the pressure (voltage) spikes until the pump explodes.
- CT Rating: 400:5A (Ratio 80:1, $N_p=1$, $N_s=80$)
- Primary Current ($I_p$): 400A RMS
- Secondary Current ($I_s$): 5A RMS
- Target Secondary Voltage ($V_s$): 2.5V RMS (for a specific analog panel meter)
Using Ohm's Law ($R = V / I$), the required burden resistance is $2.5V / 5A = \mathbf{0.5 \, \Omega}$.
Power Dissipation Check: $P = I^2 \times R = (5A)^2 \times 0.5\Omega = \mathbf{12.5W}$. You must use a burden resistor rated for at least 25W (derated 50% for safety) to prevent it from catching fire inside the panel.
What happens if you remove that $0.5\Omega$ resistor while 400A is flowing through the primary? The secondary current still attempts to push 5A, but the resistance approaches infinity. According to Faraday's Law ($V = -N \frac{d\Phi}{dt}$), the core flux ramps up to extreme saturation levels, inducing voltage spikes of 3,000V to 10,000V across the open secondary terminals. This will instantly flashover the terminal block, vaporize the fine secondary magnet wire, and pose a lethal shock hazard. Electronics-Tutorials provides an excellent breakdown of the magnetic saturation curves that cause this destructive voltage spike.
Where You Meet This in Practice (And How to Not Blow Up Your Meter)
You will encounter current transformer construction in two vastly different worlds, each with its own rules and hardware.
1. Industrial Motor Control Centers (MCCs)
In commercial panels, you will find solid-core or split-core CTs rated for 5A secondaries, built with silicon steel laminations. These feed analog ammeters, power factor correction controllers, and protective relays. The standard here is governed by IEEE C57.13, which defines accuracy classes (e.g., Class 0.5 for metering, Class C100 for protection). In these environments, the secondary wiring is often 12 AWG or 10 AWG THHN to keep the wire resistance (which adds to the total burden VA) as low as possible over long conduit runs.
2. DIY and Home Energy Monitoring
For makers building ESP32 or Arduino-based energy monitors, the industry standard is the YHDC SCT-013 series. These use cheap ferrite cores and output milliamps instead of amps. The SCT-013-000 outputs 50mA at 100A primary. Because it outputs current, you must add an external burden resistor (typically 33$\Omega$ for a 5V Arduino, or 18$\Omega$ for a 3.3V ESP32) to generate a voltage the ADC can read. Conversely, the SCT-013-030 has a built-in internal burden resistor and outputs 0-1V directly. Never use the SCT-013-030 with a multimeter in current-measuring mode; you will short out the internal burden and read zero.
FAQ: Current Transformer Construction Gotchas
Can I use a standard current transformer to measure DC or solar panel output?
No. Standard CTs rely on Faraday’s Law of Induction, which requires a changing magnetic field ($di/dt$) to induce a secondary current. DC current creates a static magnetic field, resulting in zero secondary output and eventual core saturation. To measure DC, you must use a Hall-effect sensor (like the ACS712 or a closed-loop LEM module) or a fluxgate sensor.
Why does polarity (P1/P2 and S1/S2) matter if I'm just measuring amps?
If you are only measuring raw current magnitude with an ammeter, polarity doesn't matter. However, if you are measuring real power (Watts) or power factor, the phase relationship between voltage and current is critical. If you install the CT backwards (or wire S1/S2 reversed), the microcontroller or power analyzer will see the current 180° out of phase with the voltage, resulting in a negative Watt reading and a power factor of -1.0. Always point the P1 arrow toward the load.
How do I compensate for the phase shift introduced by the CT core?
Magnetic cores introduce a slight phase angle error between the primary and secondary waveforms. In precision OpenEnergyMonitor builds, this phase shift causes errors in real power calculations at low power factors. You compensate for this in software by applying a phase calibration offset (shifting the voltage or current waveform array by a few microseconds in your ESP32 code) rather than altering the physical CT construction.
