The Physics: Voltage Across (Not Through) an Inductor
Let's clear up a common bench-top terminology mix-up right away: while hobbyists frequently search for the 'voltage through an inductor,' physically, current flows through the wire windings while voltage drops across the terminals. Understanding this distinction is critical because the voltage stress an inductor experiences is not a simple static rating like a capacitor's DC working voltage. Instead, it is governed by the rate of change of current and the core's magnetic limits.
The governing equation is V = L(di/dt). When you apply a DC voltage across an inductor, current ramps up linearly. When you suddenly interrupt that current (like a MOSFET switching off in a buck converter or a relay opening), dt approaches zero. The inductor fights this change by generating a massive reverse voltage spike—known as back-EMF or flyback voltage. If you don't manage this voltage across the component, it will arc across your switching transistor or break down the inductor's internal wire insulation.
When selecting an inductor for a switching power supply, ignore the 'maximum voltage' spec and look at the Volt-microsecond product. If your converter applies 12V across a 10µH inductor for a 5µs on-time, your V·µs is 60. If the inductor's core cannot support 60 V·µs without saturating, its inductance will collapse to near-zero, effectively creating a dead short across your power rail. Always verify the V·µs rating against your specific switching frequency and duty cycle.
Inductor Type Comparison: Matching Core to Voltage Stress
Different core materials handle magnetic flux density (and therefore voltage-time stress) differently. Picking the wrong core for your voltage profile leads to premature saturation or catastrophic thermal failure. Here is how the primary inductor constructions compare on the bench.
| Type / Construction | Tolerance | Tempco (ppm/°C) | Saturation Behavior | Typical Use Case |
|---|---|---|---|---|
| Shielded Drum (Ferrite core, epoxy/metal shield) |
±20% | -100 to -500 | Hard knee (sudden drop) | Standard buck/boost converters, LED drivers |
| Composite / Molded (Alloy powder, resin molded) |
±20% | Flat / Stable | Soft knee (gradual drop) | High transient current, automotive, tight EMI limits |
| Toroidal (Donut ferrite/iron powder, wire wound) |
±10% | Varies widely | Hard knee | High-Q EMI filters, audio crossovers, offline SMPS |
| Air Core / Ceramic (No magnetic material) |
±5% | Near Zero | None (cannot saturate) | RF matching networks, VHF chokes, high-frequency snubbers |
Decoding Physical Markings and SMD Codes
Unlike resistors, inductors do not have a single universal color-code standard, but surface-mount (SMD) power inductors generally follow a predictable 3-digit or alphanumeric marking scheme. Misreading these on a salvaged board is a fast track to blowing up a prototype.
The SMD 3-Digit System
Most shielded drum and composite SMD inductors use a system where the first two digits are the significant figures, and the third digit is the multiplier (number of zeros), expressed in microhenries (µH).
- 100 = 10 × 10^0 = 10 µH
- 101 = 10 × 10^1 = 100 µH
- 472 = 47 × 10^2 = 4700 µH (or 4.7 mH)
The 'R' Decimal System
For values under 10 µH, the letter 'R' acts as the decimal point.
- 4R7 = 4.7 µH
- R22 = 0.22 µH (220 nH)
- R10 = 0.10 µH (100 nH)
RF ceramic inductors (like the Coilcraft 0603CS series) often use the same 3-digit code but express the base unit in nanohenries (nH), not microhenries. A marking of '101' on a tiny RF inductor means 100 nH (0.1 µH), whereas on a power inductor, it means 100 µH. Always check the physical size and datasheet; a 5mm power inductor is µH, a 1.6mm RF inductor is nH.
Failure Modes: Visual Symptoms of Voltage Overstress
When an inductor is subjected to voltage spikes beyond its dielectric limits, or when it is driven into deep saturation by excessive V·µs, it fails in distinct ways. Recognizing these visual symptoms on the workbench saves hours of oscilloscope debugging.
1. Dielectric Breakdown (Insulation Failure)
The Cause: The back-EMF spike exceeds the dielectric strength of the thin enamel coating on the copper windings, causing internal arcing between adjacent turns.
Visual Symptoms: Look for a distinct 'fishy' or burning epoxy smell. On shielded SMD parts, you may see a hairline crack in the ferrite shield or discolored (brown/black) potting compound weeping from the seams. Under a microscope, the copper windings will show pitting or fused turns.
The Fix: The part is dead. You must increase the physical size of the inductor to get thicker wire enamel, or add an external RC snubber / TVS diode to clamp the flyback voltage.
2. Core Saturation (Functional Failure)
The Cause: The applied V·µs product exceeds the core's magnetic flux capacity. The inductance drops to near-zero, and the inductor acts like a straight piece of wire.
Visual Symptoms: There are no visual symptoms on the inductor itself. The inductor will look perfectly fine. However, your upstream switching MOSFET will likely be cracked, vented, or blown completely off the PCB due to the massive overcurrent event.
The Fix: Replace the MOSFET and swap the inductor for one with a higher $I_{sat}$ (saturation current) rating, or reduce the duty cycle of your switching controller.
3. Thermal Runaway (AC Loss Overload)
The Cause: High-frequency voltage ripple causes excessive eddy currents and hysteresis losses in the core, or skin-effect heating in the wire.
Visual Symptoms: The solder pads on the PCB will show signs of reflow or thermal discoloration (a dark ring on the FR4 fiberglass). The inductor's top marking might be blistered or faded from sustained 120°C+ surface temperatures.
The Fix: Switch to a composite/alloy powder core which handles high AC ripple current much better than standard ferrite.
The Decision Path: Selecting Your Inductor
Stop guessing based on whatever is in your junk bin. Use this decision tree to terminate your selection process with a concrete, proven part number.
| Your Circuit Requirement | Decision Criteria | Concrete Pick (2026 Standard) |
|---|---|---|
| Compact 1MHz+ Buck Converter (Space-constrained, high ripple current) |
Needs soft saturation knee, low profile (<2mm), and excellent thermal stability to handle high AC voltage ripple without core heating. | Coilcraft XEL4020 Series (Composite core, handles massive transient spikes without hard saturation). |
| High-Current (>15A) Step-Down (Desktop PSU, motor driver output filter) |
Requires high $I_{sat}$, low DCR (DC resistance) to minimize $I^2R$ heating, and magnetic shielding to prevent EMI from coupling into nearby logic. | Bourns SRP1265A Series (Shielded drum, carbonyl iron powder, built for high DC bias). |
| Relay / Solenoid Flyback Snubber (Low frequency, massive back-EMF spikes) |
Inductance value is secondary; the primary goal is absorbing the voltage spike. You need high breakdown voltage and robust physical construction. | Standard 1N4007 Diode (Wait—don't use an inductor here. Use a flyback diode. If you must filter the coil, use a cheap Wurth WE-SD series drum inductor). |
| RF Antenna Matching / VHF Choke (>50MHz, minimal parasitic capacitance) |
Zero core saturation risk, ultra-high Q factor, and tight tolerance (±2% or better) to keep the resonant frequency exact. | Coilcraft 0805CS Series (Ceramic air-core, wirewound, industry standard for RF). |
Safe Substitution Rules for Missing Parts
You are three hours into a weekend PCB debug, and you realize the 4.7µH shielded inductor in your BOM is out of stock. Can you substitute it? Follow these hard rules to avoid destroying your prototype.
Rule 1: Never Substitute Down on $I_{sat}$
You can safely use an inductor with a higher saturation current ($I_{sat}$) than your design requires. You must never use one with a lower $I_{sat}$. If your circuit peaks at 5A, and you substitute a 4A part, the core will saturate during transient loads, collapsing the inductance and shorting your rail. Check the datasheet's $I_{sat}$ graph, not just the $I_{rms}$ (thermal) rating.
Rule 2: Shielded vs. Unshielded Swaps
Can you swap an unshielded drum inductor for a shielded one? Yes. Shielded parts have lower radiated EMI and are generally safer in mixed-signal environments.
Can you swap a shielded part for an unshielded one? Only if your circuit has no sensitive analog front-ends (like a 16-bit ADC or audio preamp) within 15mm of the inductor. Unshielded parts leak significant magnetic flux, which will induce voltage noise directly into adjacent copper traces.
Rule 3: The Inductance Value Tolerance
For power supply filtering (buck/boost), substituting a 10µH part with a 15µH part is usually safe; it will slightly lower your output ripple voltage and push your converter's crossover frequency down, improving stability. However, for resonant converters (LLC) or RF matching networks, a 20% deviation will shift your resonant frequency entirely out of band, causing the circuit to fail to start or reflect destructive reactive power back to the source. In resonant circuits, substitute only with ±5% tolerance parts.
When in doubt, prioritize the Saturation Current ($I_{sat}$) and the Physical Footprint over the exact microhenry value. A slightly higher inductance with a massive saturation margin will almost always keep a switching regulator alive, whereas an exact inductance match with a weak core will result in a dead MOSFET on the first power-up.
For deeper magnetics theory and core material characteristics, refer to the Coilcraft Magnetics Theory documentation. To brush up on the foundational AC behavior and reactance calculations, review the All About Circuits chapter on Inductive Reactance. Finally, for specific power inductor derating curves and footprint layouts, consult the Bourns Power Inductors datasheets.
