An inductor stores energy in its inductor magnetic field when current flows through its coils. The amount of energy it can hold, and how it behaves under stress, is entirely dictated by the core material around which those coils are wound. If you select the wrong core for a switching power supply, the magnetic field will collapse into saturation, turning your inductor into a near-short circuit and instantly destroying your downstream MOSFETs. This guide breaks down the physics, core material selection, physical markings, and failure diagnostics you need to specify and troubleshoot inductors on the bench.
The Physics of the Inductor Magnetic Field
The energy stored in an inductor magnetic field is calculated using the formula E = 0.5 * L * I^2, where E is energy in Joules, L is inductance in Henries, and I is current in Amps. For a practical bench example: a standard 10mH power inductor carrying 2A of DC current stores 0.02 Joules of energy. While that sounds small, in a switching regulator operating at 500kHz, that energy is transferred 500,000 times per second, representing 10 kilowatts of instantaneous power transfer capability.
The core material amplifies this magnetic field through a property called permeability ($\mu$). However, every magnetic core has a hard physical limit known as the saturation flux density ($B_{sat}$). Once the magnetic domains in the core are fully aligned, the core saturates. At this point, the inductance drops precipitously—often to just the parasitic inductance of the wire itself (a few nanohenries). According to the All About Circuits DC textbook, this sudden drop in inductance causes the current to spike linearly, limited only by the DC resistance (DCR) of the wire and the source impedance.
Core Material Comparison: Which Type for Which Job
Choosing the right core material is a balancing act between permeability, frequency response, and saturation current. Below is a spec-sheet-table comparing the four most common core constructions you will encounter in commercial and hobbyist electronics.
| Core Type | Construction | Typical Tolerance | Tempco (ppm/°C) | Typical Use Case |
|---|---|---|---|---|
| Air Core | Coiled wire, no magnetic material | ±2% to ±5% | +50 (Copper drift) | RF/VHF circuits, high-Q filters, crossover networks |
| Ferrite (MnZn) | Sintered iron oxide/manganese/zinc | ±10% to ±20% | -2000 to -6000 | SMPS transformers, common-mode chokes (<2MHz) |
| Ferrite (NiZn) | Sintered iron oxide/nickel/zinc | ±10% to ±20% | -1000 to -3000 | EMI suppression beads, RF chokes (>2MHz to GHz) |
| Powdered Iron | Iron particles insulated and pressed | ±10% to ±15% | +100 to +350 | DC-DC buck/boost inductors, high DC bias applications |
| Laminated Steel | Stacked silicon steel sheets | ±20% to ±30% | N/A (Non-linear) | 50/60Hz line filtering, audio crossovers, mains chokes |
Selection Criteria: If you are designing a 1MHz buck converter, powdered iron or specialized ferrite (like Coilcraft's XEL series) is mandatory because MnZn ferrite will suffer massive core losses (eddy currents) at that frequency. For audio frequency filtering, laminated steel or large ferrite cores prevent saturation from high peak currents. For RF tank circuits above 50MHz, always use air core or NiZn ferrite to maintain a high Q-factor.
Decoding Inductor Markings and Safe Substitution
Unlike resistors, inductor markings are notoriously inconsistent across manufacturers. However, two dominant coding systems exist for surface mount (SMD) and through-hole parts.
Reading SMD 3-Digit and R-Codes
Most SMD power inductors use a 3-digit code where the first two digits are significant figures and the third is the multiplier (number of zeros), expressed in microhenries (µH). For example, a marking of 103 means 10 × 10³ µH, which equals 10,000 µH or 10 mH. A marking of 471 is 470 µH. When a decimal is involved, the letter 'R' replaces the decimal point: 4R7 means 4.7 µH, and R47 means 0.47 µH.
Reading Through-Hole Color Bands
Axial through-hole inductors often use the Mil-spec MIL-I-2313 color band system. It reads exactly like a 4-band resistor, but the base unit is microhenries instead of ohms. A band sequence of Brown-Black-Brown-Gold translates to 1-0-10µH-±5%, yielding a 100 µH inductor with 5% tolerance.
How to Substitute Safely
When your exact BOM part (e.g., a Würth Elektronik WE-PD 744774215) is out of stock, you can substitute safely by following these three hard rules:
- Inductance (L): Must be within ±20% of the original. Going too high slows down the control loop response; going too low increases ripple current.
- Saturation Current ($I_{sat}$): Must be greater than or equal to the original. Never substitute a lower $I_{sat}$, or the inductor magnetic field will collapse during peak load transients.
- DC Resistance (DCR): Should be as close to or lower than the original. A higher DCR will cause excessive $I^2R$ heating and drop your overall converter efficiency.
Consult the Würth Elektronik Magnetics catalog or Coilcraft Design Tools to cross-reference these specific parameters before soldering a substitute.
Failure Modes: Visual Symptoms of Saturation and Thermal Runaway
Inductors fail in distinct ways, and the physical symptoms on the PCB will tell you exactly what went wrong in the circuit.
- Thermal Runaway (Overcurrent): Visual symptom: The plastic bobbin is melted, discolored (usually brown or black), and the copper wire enamel is charred. It will smell distinctly like burning sugar. Cause: The RMS current exceeded the thermal rating ($I_{rms}$) of the wire gauge, causing $I^2R$ heating that the component could not dissipate.
- Magnetic Saturation (Core Overload): Visual symptom: The inductor looks perfectly fine, but the switching MOSFET downstream is blown wide open, or the PCB traces leading to the inductor are scorched. Cause: Peak current exceeded $I_{sat}$. The inductor magnetic field collapsed, inductance dropped to near zero, and uncontrolled current spiked through the switch. You must view this on an oscilloscope using a current probe; the current waveform will show a sharp 'shark-fin' upward slope instead of a linear ramp.
- Mechanical Fracture (Thermal Shock/Dropping): Visual symptom: A visible hairline crack through the ferrite core, or the core is split in half. Cause: Ferrite is essentially ceramic and highly brittle. Dropping the PCB or severe thermal cycling can crack it. This introduces an unintended air gap, drastically lowering the inductance and increasing EMI.
- Microphonics (The 'Singing' Inductor): Visual symptom: None, but highly audible. Cause: Magnetostriction. The magnetic field causes the core material to physically expand and contract at the switching frequency. If the switching frequency drops into the 2kHz–20kHz audio range (common in burst-mode light loads), the inductor will emit a high-pitched whine. Fix this by selecting a part with a bonded core or shifting the switching frequency above 50kHz.
FAQ: Inductor Magnetic Field Questions
Does the inductor magnetic field cause EMI in nearby traces?
Yes. An unshielded inductor (like a drum core or toroid without a metal can) radiates a significant alternating magnetic field. If you route sensitive analog traces or high-impedance feedback lines within 5mm of an unshielded power inductor, the magnetic flux will induce parasitic voltages in those traces, causing noise or control loop instability. Always use magnetically shielded inductors (like the Coilcraft MSS or MSSS series) in dense, mixed-signal PCB layouts.
How do you measure the strength of an inductor magnetic field?
You cannot easily measure the magnetic field strength (H) or flux density (B) directly on a populated PCB without specialized Gauss meters or near-field probes. Instead, engineers measure the effects of the field. You use an LCR meter to measure the inductance value at the operating frequency, and a current probe on an oscilloscope to observe the saturation point. When the current slope on the scope transitions from a straight line to a sharp exponential curve, you have visually identified the exact point where the magnetic field has saturated the core.
Why does my inductor get hot even with no DC load?
If an inductor in an AC or switching circuit heats up with zero DC load current, you are experiencing core losses, specifically hysteresis and eddy current losses. Every time the magnetic field reverses direction, the magnetic domains in the core physically flip, generating internal friction (heat). If you are using a powdered iron or standard ferrite core at a frequency higher than its design limit (e.g., running a 100kHz-rated core at 2MHz), core losses will dominate and cook the component even if the copper wire is carrying minimal current.
Can I add a second core to double the inductor magnetic field capacity?
Stacking two identical ferrite cores over the same winding will roughly double the inductance (because you doubled the cross-sectional area of the magnetic path), but it does not double the saturation current ($I_{sat}$). The saturation current is determined by the magnetic field intensity (H), which depends on the number of turns and the current, not the core volume. To increase the saturation current handling, you must either increase the physical air gap in the core, use a core material with a higher intrinsic $B_{sat}$ (like specialized Molypermalloy powder), or reduce the number of turns.
