When designing or repairing a power converter, the inductor system is not just a passive coil of wire; it is a complex magnetic energy storage device defined by its core material, winding geometry, and thermal coupling to the PCB. The direct answer to selecting the right inductor system for a DC-DC buck or boost converter hinges on three non-negotiable parameters: inductance (µH), saturation current ($I_{SAT}$), and DC resistance (DCR). If your $I_{SAT}$ is lower than your peak switch current, the core saturates, inductance collapses, and your switching FET will likely fail from overcurrent. This guide breaks down the physical construction of inductor systems, how to read their cryptic SMD markings, how they fail on the bench, and the exact framework for substituting them when your BOM is out of stock.

Inductor System Types: Which Core for Which Job?

The core material of an inductor system dictates its energy storage capacity, loss profile, and behavior under temperature and DC bias. According to Coilcraft's magnetics basics, selecting the wrong core material for a high-frequency switcher will result in excessive core losses (hysteresis and eddy currents) that cook the component from the inside out. Below is a comparison of the four primary inductor system constructions used in modern power electronics.

Core Material / Construction Typical Tolerance Tempco / Temp Stability Saturation Behavior Typical Use Case
Ferrite (Unshielded)
Open magnetic path, bobbin wound		 ±10% to ±20% Poor above 100°C (Curie temp limits) Hard saturation (sharp inductance drop) Low-cost, low-frequency filtering; non-critical power rails where EMI is not a concern.
Ferrite (Shielded/Molded)
Ferrite powder mixed with epoxy/binder		 ±20% to ±30% Moderate; stable to ~105°C Soft saturation (gradual roll-off) General-purpose DC-DC buck converters (e.g., Wurth WE-LQS series) where moderate EMI shielding is needed.
Powdered Iron
Distributed air gap via insulated particles		 ±10% to ±15% Excellent; stable up to 125°C+ Very soft saturation High-current, continuous conduction mode (CCM) PFC chokes and high-temperature automotive environments.
Metal Alloy (Composite)
Carbonyl iron or alloy powder in resin		 ±20% Exceptional; stable to 125°C-155°C Extremely soft, predictable roll-off High-frequency (>1MHz), high-current density systems (e.g., Coilcraft XEL, TDK SPM). Very low EMI.

Selection Criteria: Choose metal alloy composite cores for modern, high-frequency point-of-load (POL) regulators where board space is tight and EMI must be minimized. Stick to gapped ferrite or powdered iron for high-current, lower-frequency applications where core volume is less constrained.

Decoding Inductor Markings and Part Codes

Unlike resistors and capacitors, SMD power inductors rarely follow a single universal color or digit standard, but the vast majority of modern shielded and composite inductors use a 3-character alphanumeric code stamped on the top shield or epoxy.

  • The 'R' Decimal Indicator: If you see 4R7, the 'R' acts as the decimal point. This means 4.7µH. Similarly, R47 means 0.47µH, and R10 means 0.10µH.
  • The 3-Digit EIA Code: For values of 10µH and above, manufacturers use two significant digits followed by a multiplier (number of zeros). 100 means 10 × 10⁰ = 10µH. 471 means 47 × 10¹ = 470µH. 102 means 10 × 10² = 1000µH (1mH).
  • Orientation Dots: Many shielded inductors feature a small white dot or an underline next to the value code. While standard inductors are non-polarized, this dot indicates the physical start of the winding. In a standard buck converter, this doesn't matter. However, in an inductor system used for a SEPIC, flyback, or coupled-inductor topology, the dot denotes phase polarity. Wiring it backward will cause catastrophic switch node ringing or controller malfunction.
Bench Warning: Never rely solely on visual markings for critical repair. The top stamping on a 10µH and a 100µH molded inductor can look identical if the '0' multiplier is worn off or obscured by flux residue. Always verify with an LCR meter set to 100kHz before soldering.

Failure Modes and Visual Symptoms on the Bench

Inductors are generally robust, but when pushed beyond their thermal or magnetic limits, they fail in highly specific ways. All About Circuits notes that core material limitations often manifest as physical damage long before the silicon switching IC fails.

1. Core Saturation and Thermal Runaway
Symptom: Melted solder fillets on the inductor pads, discolored PCB substrate (browning), and a component that is too hot to touch.
Physics: When peak current exceeds $I_{SAT}$, the core's permeability drops to near that of air. The inductor essentially becomes a low-value resistor (just the DCR of the copper wire). The massive $I^2R$ heating melts the solder joints, often lifting the pad off the PCB.

2. Inter-Turn Winding Short
Symptom: Bulging or cracked epoxy on the top shield, inductance reads near 0µH on an LCR meter, but DCR reads slightly lower than the datasheet spec.
Physics: High voltage spikes (dv/dt) on the switch node break down the thin enamel insulation between adjacent copper windings. The shorted turns act as a secondary shorted transformer coil, generating massive localized eddy currents that cook the epoxy from the inside.

3. Mechanical Shield Cracking
Symptom: Visible hairline fractures running vertically down the sides of a ferrite shielded drum core.
Physics: Ferrite is essentially ceramic. If the PCB flexes during mechanical assembly, or if the inductor is subjected to severe drop shocks, the brittle ferrite shield cracks. This alters the magnetic air gap, causing the inductance value to drop unpredictably and increasing radiated EMI.

Safe Substitution: What to Do When the Exact Part is Missing

Supply chain shortages frequently force engineers to substitute magnetics. You cannot simply swap parts based on the µH stamping. Use this strict decision framework to safely substitute an inductor system:

  1. Inductance Match (±20%): For a standard voltage-mode or current-mode buck converter, a ±20% variance in inductance is acceptable; it will slightly alter your ripple current and loop crossover frequency. For resonant topologies (LLC) or precision filters, you must match within ±5%.
  2. Saturation Current ($I_{SAT}$) - NEVER COMPROMISE: Your substitute's $I_{SAT}$ must be greater than or equal to the original part. Rule of thumb: $I_{SAT}$ must be at least 1.3 × $I_{OUT(MAX)}$ for continuous conduction mode. If you substitute a part with a lower $I_{SAT}$, you risk destroying the downstream MOSFET.
  3. DC Resistance (DCR): You can safely substitute a part with a lower DCR (it will run cooler). If you must use a part with a higher DCR, calculate the new thermal rise. A jump from 10mΩ to 30mΩ at 5A RMS increases copper loss from 0.25W to 0.75W, which may exceed the thermal limits of a small 4x4mm SMD package.
  4. Shielding and EMI: Substituting an unshielded drum core for a shielded composite part is a safe electrical upgrade, but it may physically interfere with a nearby RF shield can due to height differences. Conversely, swapping a shielded part for an unshielded one to save pennies will likely cause your design to fail FCC/CE radiated emissions testing.

Inductor System FAQs

Can I use two smaller inductors in parallel to build a larger inductor system?

Electrically, placing two identical inductors in parallel halves the total inductance ($L_{total} = L / 2$) and doubles the current handling capability, assuming perfect current sharing. However, in practice, this is highly problematic due to mutual magnetic coupling. If the two inductors are placed close together on the PCB, their magnetic fields will interact. Depending on their physical orientation, they can either couple constructively (increasing inductance and risking saturation) or destructively (canceling flux and dropping inductance to near zero). If you must parallel them for current sharing, use shielded composite inductors, space them at least one component-width apart, and orient their magnetic axes orthogonally (at 90-degree angles) to minimize mutual inductance.

Why does my inductor system whine or squeal under light loads?

This audible noise is caused by magnetostriction. When the magnetic domains in the ferrite or powdered iron core align with the alternating magnetic field, the core material physically expands and contracts at the switching frequency. Under heavy loads, the PWM switching frequency is typically 500kHz to 2MHz, which is well above human hearing. However, under light loads, many modern DC-DC controllers enter 'pulse-skipping' or 'burst mode' to save quiescent current. This drops the effective switching frequency into the 2kHz to 20kHz audible range. To fix this, look for inductors with a monolithic, fully molded composite construction (like metal alloy types), which mechanically damp the core vibrations far better than loose ferrite drums wrapped in tape.

How do I measure the saturation current of an unknown inductor on my bench?

You cannot measure $I_{SAT}$ with a standard multimeter or a basic LCR meter. $I_{SAT}$ is defined as the DC bias current at which the inductance drops by a specific percentage (usually 20% or 30%, depending on the manufacturer). To measure this on the bench, you need an LCR meter equipped with a DC bias current source accessory. You sweep the DC current from 0A up to the component's thermal limit while logging the inductance. The point where the logged inductance curve drops 30% from its 0A baseline is your $I_{SAT}$. If you lack a DC bias LCR meter, you can build a simple curve tracer using a power supply, a power MOSFET, a low-value current shunt resistor (e.g., 10mΩ), and an oscilloscope to monitor the di/dt slope across the inductor while pulsing the MOSFET.