The voltage across an inductor is not determined by the absolute current flowing through it, but strictly by the rate of change of that current. In a steady DC state, an ideal inductor acts as a short circuit (0V drop), while a real inductor drops voltage equal to $I \times DCR$ (DC Resistance). However, during switching transients—like a MOSFET turning off—the collapsing magnetic field forces the inductor to generate whatever voltage is necessary to keep current flowing, often resulting in destructive flyback spikes.
This deep-dive covers the exact physics of inductor voltage, how to select the right core material for your application, how to decode physical markings, and the failure modes you will encounter on the bench.
The Core Physics: Calculating Voltage Across an Inductor
The fundamental governing equation for inductor voltage is:
$V_L = L \frac{di}{dt}$
Where $V_L$ is the voltage across the inductor, $L$ is the inductance in Henries, and $\frac{di}{dt}$ is the rate of current change in Amperes per second.
Worked Numeric Example: Buck Converter Switching Node
Imagine you are designing a buck converter stepping 12V down to 3.3V using a 4.7µH metal-alloy inductor. When the high-side MOSFET turns on, the voltage applied across the inductor is $V_{in} - V_{out} = 12V - 3.3V = 8.7V$.
Using the rearranged formula $\frac{di}{dt} = \frac{V}{L}$, the current ramps up at a rate of:
$\frac{8.7V}{4.7 \times 10^{-6}H} \approx 1.85 \times 10^6 \text{ A/s}$ (or 1.85 A/µs).
If the FET stays on for 2µs, the inductor current ramps up by $1.85 \times 2 = 3.7A$. This predictable ramp is how switch-mode power supplies transfer energy.
The Flyback Transient (When the Switch Opens)
When the MOSFET turns off, the current path is broken. Mathematically, $dt$ approaches zero, meaning $\frac{di}{dt}$ approaches infinity. To maintain current continuity, the inductor's voltage will spike until it finds a path—often avalanching the MOSFET's drain-source junction or arcing across a relay contact. This is why flyback diodes or RC snubber networks are mandatory across inductive loads.
Inductor Construction Types and Selection Matrix
Choosing the wrong core material will result in premature saturation, excessive EMI, or thermal failure. The table below maps core constructions to their real-world electrical characteristics and ideal use cases.
| Core Material / Type | Typical Tolerance | Tempco (ppm/°C) | Saturation Behavior | DCR & EMI Traits | Typical Use Case |
|---|---|---|---|---|---|
| Ferrite Drum (Shielded) | ±20% | -2000 to +5000 | Hard saturation (sharp inductance drop) | Medium DCR, excellent magnetic shielding | DC-DC step-down converters, general power rails |
| Powdered Iron (Toroid) | ±10% | +50 to +350 | Soft saturation (gradual inductance roll-off) | High core loss at high freq, zero EMI radiation | PFC chokes, audio crossover filters, high-current AC |
| Air Core | ±2% to ±5% | ~0 (Copper only) | Cannot saturate (no magnetic core) | Very low inductance, high DCR for large values | RF tuning, VHF/UHF antennas, high-end audio crossovers |
| Metal Alloy Composite (Molded) | ±20% | -1000 to +1000 | Very soft saturation, handles high peak currents | Extremely low DCR, low profile, low EMI | CPU/GPU VRMs, automotive high-current rails, LiPo BMS |
Selection Rule of Thumb: If your circuit experiences high peak-to-average current ratios (like a buck converter with discontinuous conduction mode), choose Metal Alloy or Powdered Iron for their soft saturation curves. If you need strict magnetic shielding to prevent noise coupling into a nearby 2.4GHz antenna, use a Shielded Ferrite Drum.
Decoding Inductor Markings and Safe Substitution
Unlike resistors, inductor markings are notoriously inconsistent across manufacturers. Here is how to read the physical part and how to substitute safely when your exact BOM part is out of stock.
Reading SMD and Through-Hole Codes
- SMD 3-Digit Code: The first two digits are the significant figures, and the third is the multiplier (number of zeros) in microhenries (µH). A marking of
471means 47 × 10¹ = 470µH. A marking of4R7uses 'R' as a decimal point, meaning 4.7µH. - SMD 4-Digit Code: Used for tighter tolerances.
1001means 100 × 10¹ = 1000µH (or 1mH). - Through-Hole Color Bands: Read similarly to resistors, but the unit is microhenries. A band sequence of Brown-Black-Brown-Silver translates to 1-0-×10 = 100µH with a ±10% tolerance. (Note: Some manufacturers use a double-width first band to indicate an RF inductor in nanohenries).
When substituting an inductor, matching the inductance (L) is not enough. You must verify both $I_{rms}$ (thermal current limit based on DCR heating) and $I_{sat}$ (saturation current where inductance drops by 20-30%). Never substitute a part with a lower $I_{sat}$. You can safely substitute a higher $I_{sat}$ part, provided it physically fits the footprint and its DCR is not so low that it causes control-loop instability in certain hysteretic regulators.
Failure Modes: When Voltage Spikes Destroy Components
Inductors rarely fail silently. When the voltage across an inductor exceeds its dielectric limits, or when current exceeds its magnetic limits, the physical symptoms are distinct.
1. Core Saturation (Invisible Inductor Failure)
The Mechanism: The peak current exceeds $I_{sat}$. The magnetic core saturates, and the inductor's permeability drops to that of air. The inductance value plummets, causing $\frac{di}{dt}$ to spike massively.
Visual Symptom: The inductor looks perfectly fine. However, the driving MOSFET will be shattered, or the PCB trace leading to the drain pin will be vaporized due to the massive overcurrent event.
Fix: Increase the inductor's physical size, switch to a metal-alloy composite core, or add a cycle-by-cycle current limit to your controller.
2. Inter-Winding Insulation Breakdown
The Mechanism: High flyback voltages (often hundreds of volts in relay coils or flyback transformers) exceed the dielectric strength of the thin enamel coating on the copper wire. An arc jumps between adjacent windings, creating a shorted turn.
Visual Symptom: Discolored or melted heat-shrink tubing, a distinct burnt-ozone smell, and bulging on the side of the inductor. If you measure it with a multimeter, the DCR will be significantly lower than the datasheet specification.
Fix: Replace the component and add an RC snubber or TVS diode to clamp the flyback voltage below the enamel's breakdown threshold (typically 500V-1kV for standard magnet wire).
3. Mechanical Pad Fracture (SMD)
The Mechanism: Thermal cycling or board flexion cracks the solder joint or the ferrite core itself. This is common in large, unshielded drum-core inductors placed near board mounting holes.
Visual Symptom: A visible hairline crack in the ferrite drum, or the component lifts off the pad. Intermittent operation when the board is pressed.
Fix: Use a smaller footprint, switch to a molded metal-alloy inductor (which are mechanically robust and resist cracking), or add underfill/potting compound.
Practical Bench Measurement and Snubber Design
Measuring the voltage across an inductor in a switching circuit is one of the most common ways junior engineers destroy their oscilloscopes.
The Ground-Clip Death Trap
If you are measuring a high-side inductor in a non-isolated buck converter, the switching node is referenced to ground through the low-side FET. If you clip your oscilloscope's ground lead to the switching node, you are directly shorting that node to earth ground through the scope's power cord. This will instantly vaporize the ground clip, destroy the probe, and likely blow up your board's low-side FET.
The Correct Method: Use a high-voltage differential probe to measure across the inductor. If you do not have a differential probe, you must use the "A minus B" math function on your scope with two matched single-ended probes, ensuring neither probe's ground clip is connected to the circuit (float the grounds using an isolation transformer on the scope, or use battery-powered handheld scopes).
Designing the Snubber
If your scope capture shows a flyback spike ringing at 50MHz with a peak of 80V on a 12V rail, you need an RC snubber. 1. Measure the ringing frequency ($f_{ring}$) with no snubber. 2. Add a capacitor ($C_{add}$) across the inductor until the ringing frequency drops by half. This means $C_{add}$ is equal to the parasitic capacitance ($C_{par}$) of the circuit. 3. Calculate the characteristic impedance: $Z_0 = \frac{1}{2 \pi f_{ring} C_{par}}$. 4. Set your snubber resistor $R_s = Z_0$, and your snubber capacitor $C_s = 4 \times C_{par}$. This critically damps the LC tank formed by the inductor and parasitic capacitance, safely clamping the voltage across the inductor without excessive steady-state power loss.
