Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61185 — Single Current Pulse Method for Magnetic Core Magnetization Characterization
In modern power electronics and switch-mode power supply (SMPS) design, the magnetic core characteristics of inductive components directly dictate converter efficiency, power density, and electromagnetic compatibility performance. IEC 61185 establishes a standardized single current pulse method for measuring core magnetization characteristics, offering engineers a reliable path to accurately acquire critical parameters such as the B-H loop, saturation flux density (B_s), remanence (B_r), and coercivity (H_c), without inducing significant core self-heating. This article provides an in-depth technical examination of the standard’s test principles, implementation schemes, and engineering best practices.
Core Technical Insight: Unlike conventional AC excitation methods that sweep the B-H loop continuously, IEC 61185 employs a single current pulse to complete the entire magnetization sweep from the demagnetized state through saturation. This fundamentally eliminates self-heating errors that plague traditional measurements, making it especially valuable for characterizing low-loss ferrite cores at high flux densities.
1. Test Principle and Theoretical Foundation of the Single Current Pulse Method
1.1 The Physics of Magnetization Under Pulsed Excitation
Magnetization of a ferromagnetic material is fundamentally a process of magnetic domain wall motion and rotation under an applied external field. Conventional AC excitation repeatedly sweeps the B-H loop via a continuous alternating field. While this yields a complete hysteresis loop, the accumulated core losses — both eddy-current and hysteresis — cause significant temperature rise, which in turn alters the intrinsic magnetic properties. This is particularly problematic for Mn-Zn and Ni-Zn ferrites, whose Curie temperature and initial permeability are strongly temperature-dependent.
IEC 61185’s single current pulse method circumvents this by applying a single, controlled-amplitude, controlled-duration current pulse to the core under test, completing the full magnetization sweep from the initial state to deep saturation within a single excitation event. The pulse width typically ranges from 10 microseconds to 100 milliseconds, depending on the core material and target measurement parameters.
B(t) = (1 / N · Ae) ∫ [V2(t) − (Rm · i2(t))] dt
Where N is the number of secondary turns, Ae is the effective cross-sectional area of the core, V2(t) is the induced secondary voltage, and Rm · i2(t) represents the resistive compensation term for the secondary circuit.
1.2 Core Architecture of the Test System
Per IEC 61185, a compliant test system comprises: a programmable pulse generator capable of delivering controlled amplitude and width current pulses; primary (excitation) and secondary (sense) windings on the core under test; and a high-bandwidth, high-resolution digital acquisition system for simultaneous voltage and current waveform capture. In practice, the primary current is sensed through a low-inductance precision shunt resistor, while the secondary open-circuit voltage directly reflects the flux rate of change dΦ/dt.
Engineering Caution: Single-pulse measurements demand exceptional sampling synchronization and noise immunity. The rising-edge slew rate (di/dt) of the pulse current directly determines the signal-to-noise ratio of the induced secondary voltage. Improper trigger timing or ground-loop interference can cause severe distortion in the reconstructed B-H loop. Differential probes and shielded twisted-pair signal routing are strongly recommended.
2. B-H Curve Reconstruction Algorithms and Key Parameter Extraction
2.1 Computing Magnetic Field Strength and Flux Density
The magnetic field strength H is derived directly from the primary current i1(t) and the core’s effective magnetic path length le:
H(t) = N1 · i1(t) / le
Obtaining the magnetic flux density B requires numerical integration of the induced secondary voltage. IEC 61185 specifically mandates baseline correction of the induced voltage signal to eliminate integration drift — the single most critical signal-processing step in the entire measurement. Widely adopted correction techniques include pre-pulse baseline averaging and linear detrending algorithms.
2.2 Key Magnetic Parameters and Their Engineering Relevance
| Parameter | Symbol | Definition | Engineering Significance |
|---|---|---|---|
| Saturation Flux Density | Bs | B value where magnetization curve flattens | Sets max flux swing for transformers; directly limits power density |
| Remanence | Br | Residual flux density after current returns to zero | Affects core reset conditions and dead-time design in single-ended converters |
| Coercivity | Hc | Reverse field needed to reduce B to zero | Reflects core loss level; lower Hc means lower hysteresis loss |
| Amplitude Permeability | μa | Bmax / (μ0 · Hmax) | Large-signal effective permeability; used for inductance calculation |
| Pulse Permeability | μp | Differential permeability at pulse peak | Directly applicable to magnetizing inductance estimation in pulse transformers |
Design Insight: In LLC resonant converter design, magnetization curves measured per IEC 61185 enable precise magnetic integration optimization. For example, single-pulse measurements on PC95 grade ferrite reveal that Bs at 100°C drops to 72% of its 25°C value — a direct indication that elevated operating temperatures demand either larger core cross-sections or fewer turns to avoid saturation.
3. Engineering Practice and Design Considerations
3.1 Matching Pulse Width to Material Characteristics
The choice of pulse width profoundly affects measurement outcomes. For high-frequency power ferrites (e.g., 3F4, N49 grades), whose domain-wall switching times lie in the sub-microsecond range, pulse widths of 10–50 μs suffice for complete magnetization sweeps. For nanocrystalline or amorphous cores, whose high-permeability domain structures are more complex, 1–10 ms pulse widths are required to ensure full magnetization.
Excessively short pulses yield incomplete magnetization and artificially low Bs readings. Conversely, overly long pulses risk introducing winding losses from skin and proximity effects. IEC 61185 recommends performing a preliminary sweep with a wide pulse on unknown materials, then optimizing the pulse parameters based on the observed knee-point location.
3.2 Winding Design and Demagnetization Strategy
The choice of test winding turns involves a trade-off between signal-to-noise ratio and stray inductance. IEC 61185 recommends N1 = 3–10 turns for the primary and N2 = 10–50 turns for the secondary. Higher secondary turns amplify the induced voltage signal but also increase inter-winding capacitance, potentially introducing resonant ringing.
After each pulse test, residual remanent flux remains in the core. If not properly removed, it compromises the initial-state consistency of subsequent measurements. A demagnetization sequence with exponentially decaying alternating excitation (1–10 kHz) should be applied between single-pulse tests to restore the core to a defined neutral magnetic state. Incomplete demagnetization can introduce errors exceeding ±15% in Br and Hc values.
Common Pitfall: Many practitioners mistakenly equate the single-pulse test with a DC magnetization curve measurement. It is critical to understand that the B-H trajectory obtained per IEC 61185 is not a static DC curve but a dynamic magnetization curve recorded under specific pulse-width and di/dt conditions. For loss-sensitive magnetic materials, the discrepancy between dynamic and static Bs values can reach 8–12%. Design engineers must select test conditions that match their actual operating frequency.
3.3 Noise Handling and Integration Correction in Data Analysis
Numerical integration of the induced voltage signal is extremely sensitive to low-frequency noise and DC offset. While IEC 61185 does not mandate a specific algorithm, three approaches are widely adopted in engineering practice:
- Baseline correction: 200–500 pre-pulse samples are averaged to compute the DC offset, which is then subtracted from the full waveform.
- Polynomial detrending: A low-order polynomial is fitted to the integrated B signal to remove linear or quadratic drift components.
- Symmetrization algorithm: Using the physical symmetry of the magnetization curve, the integrated result is constrained and corrected toward the origin.
A digitizer with at least 16-bit resolution and a sampling rate no lower than 10 MS/s is the minimum hardware prerequisite for obtaining high-quality results.
3.4 Temperature Characterization Protocol
Given the strong temperature dependence of ferrite properties, IEC 61185-compliant characterization should be performed across the intended operating temperature range. For power transformer design, magnetization curves at a minimum of three temperature points — 25°C, 60°C, and 100°C — should be acquired. The pulse amplitude should be adjusted at each temperature to ensure that true saturation is reached, as Bs decreases with rising temperature while the required H for saturation remains relatively stable.
An often-overlooked detail is thermal stabilization time: the core must be held at the target temperature for sufficient duration (typically 15–30 minutes for ferrite cores above 20 mm diameter) to ensure uniform temperature distribution throughout the magnetic volume before pulse testing commences.
4. Frequently Asked Questions
Q1: Can the single current pulse method completely replace conventional AC magnetization measurement?
No. The single-pulse method excels at avoiding self-heating effects and is ideal for high-B characterization, but conventional AC methods remain indispensable for high-frequency loss measurement (e.g., Steinmetz parameter extraction). The two techniques are complementary; best engineering practice employs both during the design validation phase.
No. The single-pulse method excels at avoiding self-heating effects and is ideal for high-B characterization, but conventional AC methods remain indispensable for high-frequency loss measurement (e.g., Steinmetz parameter extraction). The two techniques are complementary; best engineering practice employs both during the design validation phase.
Q2: How significantly does ambient temperature affect test results?
Very significantly. For Mn-Zn ferrites, Bs at 100°C can drop to 65–75% of its room-temperature value. IEC 61185 recommends conducting tests inside a temperature-controlled chamber with continuous core temperature monitoring. For production-quality power transformer designs, magnetization curves at 25°C, 60°C, and 100°C represent the minimum acceptable characterization set.
Very significantly. For Mn-Zn ferrites, Bs at 100°C can drop to 65–75% of its room-temperature value. IEC 61185 recommends conducting tests inside a temperature-controlled chamber with continuous core temperature monitoring. For production-quality power transformer designs, magnetization curves at 25°C, 60°C, and 100°C represent the minimum acceptable characterization set.
Q3: How does the test method differ between toroidal and EE/EI core geometries?
Toroidal cores, with their closed magnetic path and minimal leakage flux, yield the most accurate results with simple direct winding. EE and EI cores must account for air-gap effects — even well-lapped mating surfaces introduce an effective gap that reduces apparent permeability. For EE cores, tightly wound center-post windings with controlled clamping pressure are recommended to maintain consistent gap conditions across measurements.
Toroidal cores, with their closed magnetic path and minimal leakage flux, yield the most accurate results with simple direct winding. EE and EI cores must account for air-gap effects — even well-lapped mating surfaces introduce an effective gap that reduces apparent permeability. For EE cores, tightly wound center-post windings with controlled clamping pressure are recommended to maintain consistent gap conditions across measurements.
Q4: What should be done when single-pulse Bs values disagree with datasheet specifications?
First, verify that the pulse-width conditions are comparable. Datasheets typically quote Bs under AC excitation at 10 kHz or higher, where self-heating reduces the measured value. The single-pulse method, free of self-heating, typically yields values 5–10% higher. Additionally, batch-to-batch Bs variation of ±8% is normal for ferrite cores. Maintaining an in-house database of sampled measurements per incoming batch is strongly advised for critical designs.
First, verify that the pulse-width conditions are comparable. Datasheets typically quote Bs under AC excitation at 10 kHz or higher, where self-heating reduces the measured value. The single-pulse method, free of self-heating, typically yields values 5–10% higher. Additionally, batch-to-batch Bs variation of ±8% is normal for ferrite cores. Maintaining an in-house database of sampled measurements per incoming batch is strongly advised for critical designs.
