The voltage on an inductor is not a fixed, static value like the voltage across a resistor. Instead, it is generated dynamically by the rate of change of current flowing through it, defined by the fundamental formula V = L(di/dt). When current through an inductor is interrupted abruptly, this voltage can spike to hundreds or even thousands of volts—a phenomenon known as inductive kickback or flyback voltage. Understanding this behavior is the difference between a reliable power supply and a workbench littered with shattered silicon.
The Core Physics: Why Voltage Spikes on an Inductor
To understand the voltage on an inductor, you have to look at energy storage. An inductor stores energy in its magnetic field ($E = \frac{1}{2}LI^2$). When you force current through it, the magnetic field builds. When you open the switch (like a MOSFET turning off), the magnetic field collapses. The inductor will generate whatever voltage is necessary to keep the current flowing at that exact instant.
Think of it like water hammer in plumbing. If water is flowing through a pipe and you slam a valve shut, the kinetic energy of the water creates a massive pressure spike that can rupture the pipe. The inductor is the water, the current is the flow rate, and the voltage spike is the pressure spike.
V = 0.001 H × (2 A / 0.000001 s) = 2,000 Volts.Even from a 5V microcontroller circuit, a fast-switching inductive load can easily generate enough voltage to arc across switch contacts or punch through a transistor's gate oxide.
Bench War Story: Blowing a MOSFET with Flyback Voltage
Abstract formulas are fine, but let us look at a real-world scenario where ignoring the voltage on an inductor destroyed a prototype.
The Setup
I was building a 12V to 24V DC-DC boost converter to drive a stepper motor. The design used an IRF3205 N-channel MOSFET (rated for 55V $V_{DSS}$) and a 47µH molded power inductor (Bourns SRP1265A series). The switching frequency was set to 100kHz via a TL494 PWM controller, and I used a through-hole 1N5822 Schottky diode for the flyback path.
The Numbers
At a 5A peak inductor current, the MOSFET turns off. The current must commutate from the MOSFET to the Schottky diode. The $di/dt$ is massive. The theoretical flyback voltage is clamped by the output voltage (24V) plus the diode forward voltage (~0.5V), meaning the MOSFET drain should only see about 24.5V plus a small ringing margin. Well within the 55V limit.
The Outcome
On the third power-up, the IRF3205 failed violently, shorting the gate to the drain and taking the TL494 driver stage with it. Hooking up an oscilloscope with a high-voltage differential probe revealed the culprit: a 120V spike on the drain pin lasting for about 40 nanoseconds before settling.
What Went Wrong
The 1N5822 diode was too slow, and worse, its long through-hole leads introduced parasitic inductance. When the MOSFET turned off, the parasitic inductance of the diode leads delayed the clamping action. During that 40ns delay, the inductor's flyback voltage spiked to 120V, exceeding the 55V avalanche rating of the IRF3205. The fix was twofold: I replaced the through-hole diode with a surface-mount SS34 Schottky (minimizing lead inductance) and added an RC snubber (100Ω and 1nF) across the MOSFET drain and source to dampen the high-frequency ringing.
Inductor Types and Selection Criteria
Not all inductors handle voltage spikes, saturation, and high frequencies equally. Choosing the right type for the job dictates your circuit's reliability. Below is a comparison of the most common inductor constructions you will encounter.
| Inductor Type | Construction | Typical Tolerance | Tempco (ppm/°C) | Typical Use Case |
|---|---|---|---|---|
| Molded Power (e.g., SRP1265A) | Wire wound around a composite metal alloy core, encased in epoxy | ±20% | +100 to +300 | Switch-mode power supplies (SMPS), boost/buck converters |
| Toroidal (e.g., Amidon T-50-2) | Wire wound around a powdered iron or ferrite donut ring | ±10% to ±20% | +20 to +100 | RF filters, EMI chokes, audio crossovers |
| Multilayer Ceramic (e.g., Taiyo Yuden LBMF) | Ferrite paste and conductive coils printed and sintered in layers | ±10% to ±20% | +200 to +500 | High-frequency RF matching, low-current signal filtering |
| Air Core | Wire wound on a non-magnetic plastic or ceramic former (or self-supporting) | ±5% to ±10% | ~0 (Negligible) | VHF/UHF radio tuning, high-current applications where saturation is unacceptable |
Which type for which job? If you are building a DC-DC converter handling amps of current, always choose a shielded molded power inductor. The shielding prevents the alternating magnetic field from inducing noise into nearby traces. For RF oscillators where frequency stability across temperature is critical, use an air core or a specific temperature-stable toroid (like Amidon Type-6 powdered iron). Never use multilayer ceramic inductors for power conversion; their saturation current is usually measured in milliamps.
Decoding Inductor Markings and Safe Substitution
When you are scavenging parts or reading a schematic, you need to know what the physical markings mean. Most surface-mount and radial leaded inductors use a standard three-digit EIA code, similar to resistors, but the base unit is microhenries (µH), not ohms.
- 470 = 47 × 10^0 = 47µH
- 101 = 10 × 10^1 = 100µH
- 102 = 10 × 10^2 = 1000µH (or 1mH)
For values under 10µH, manufacturers use an 'R' to denote the decimal point. A marking of 4R7 means 4.7µH, and R22 means 0.22µH.
How to Substitute Safely
If the exact part is out of stock, substituting an inductor is far more dangerous than substituting a resistor or capacitor. You must match or exceed four parameters:
- Inductance (L): Keep it within 20% of the original. Too low, and your ripple current increases; too high, and your control loop may become unstable.
- Saturation Current ($I_{sat}$): The current at which inductance drops by 20% to 30%. Never substitute a part with a lower $I_{sat}$. If the inductor saturates, it effectively becomes a piece of wire, causing massive current spikes that will destroy your switching transistor.
- RMS Current ($I_{rms}$): The current limit dictated by wire heating (DCR). Substituting a lower $I_{rms}$ part will result in a melted component.
- DC Resistance (DCR): Lower is generally better for efficiency, but drastically lowering DCR can sometimes alter the Q-factor in tuned RF circuits.
Failure Modes: Visual Symptoms and Bench Diagnostics
Inductors are generally robust, but they do fail. When they do, the voltage on an inductor can behave erratically, causing downstream chaos. Here is how to identify failure modes on the bench.
1. Core Saturation (Thermal Runaway)
What happens: The current exceeds $I_{sat}$, permeability drops, and the inductor acts like a short circuit during the peak of the switching cycle. The MOSFET overheats and fails.
Visual Symptoms: The inductor itself may look perfectly fine, but the switching transistor will be charred or cracked. The inductor core may be hot enough to melt solder or discolor the PCB.
2. Insulation Breakdown (Inter-winding Short)
What happens: High voltage spikes (flyback) exceed the dielectric strength of the thin enamel coating on the copper wire. Turns short together, drastically reducing the total inductance.
Visual Symptoms: Look for a 'flashover' mark—a tiny black pinprick or carbon track on the outer wrapping or epoxy coating. You may smell burning ozone or melting plastic. An LCR meter will show an inductance value significantly lower than the marking indicates.
3. Open Circuit (Vibration or Thermal Fatigue)
What happens: Repeated thermal cycling (expansion and contraction) or mechanical vibration breaks the copper wire, usually right at the termination point where the wire meets the solder pad.
Visual Symptoms: No visible damage to the body. However, a multimeter continuity test will read 'OL' (open loop). In SMD inductors, you may see a microscopic crack in the solder fillet under 10x magnification.
Taming the Spike: Snubbers and Flyback Diodes
You cannot stop an inductor from generating voltage when current changes, but you can control where that voltage goes. Managing the voltage on an inductor is entirely about providing a safe, low-impedance path for the stored energy to dissipate or transfer.
Flyback Diodes: For DC relays, solenoids, and simple motor drives, a reverse-biased diode placed in parallel with the inductor is mandatory. When the switch opens, the inductor's voltage reverses polarity, forward-biasing the diode and allowing the current to circulate until it decays. For high-speed switching (like SMPS), standard 1N4007 rectifiers are useless due to their slow reverse recovery time ($t_{rr}$). You must use ultra-fast recovery diodes (UF4007) or Schottky diodes (SS34, MBRS140) which turn on almost instantly.
RC Snubbers: When a diode alone is not enough to stop high-frequency ringing caused by parasitic capacitance and inductance, an RC snubber is used. A resistor and capacitor in series are placed across the switch or the inductor. The capacitor absorbs the high-frequency voltage spike, while the resistor dissipates the energy as heat and damps the resonance. As noted in Analog Devices' application notes, tuning a snubber requires measuring the ringing frequency on an oscilloscope and calculating the characteristic impedance to select the correct R and C values.
Ultimately, respecting the physics of $V = L(di/dt)$ and selecting components based on their saturation limits and parasitic properties will save you from the most common, and most destructive, mistakes in power electronics design.
