An inductor is just a coil of wire until you wrap it around a core. That core material dictates everything from saturation current and thermal limits to Q-factor and electromagnetic interference (EMI). If you treat all inductors as interchangeable black boxes that just "store energy in a magnetic field," you will eventually smoke a switching regulator or detune an RF filter. This guide cuts through the abstract physics and gives you the bench-level criteria for selecting, reading, and substituting the right inductor for your circuit.
The Direct Answer: Matching Inductor Types to the Job
The right inductor depends entirely on your operating frequency and current profile. Power conversion demands high saturation currents and low DC resistance (DCR), while RF circuits demand high Q-factors and tight tolerances. Here is the selection matrix for the four most common types of inductors you will encounter on a modern PCB.
| Inductor Type | Construction & Core | Typical Tolerance | Tempco (ppm/°C) | Primary Application |
|---|---|---|---|---|
| Air Core | Enamel wire wound on ceramic/plastic former; µr = 1 | ±2% to ±5% | +20 to +50 | VHF/UHF RF, high-Q filters, crossover networks |
| Shielded Ferrite | MnZn or NiZn powder encapsulated in a magnetic shield can | ±10% to ±20% | Non-linear (varies with temp) | DC-DC buck/boost converters, EMI-sensitive analog rails |
| Unshielded Ferrite (Drum) | Wire wound on a ferrite bobbin, exposed magnetic path | ±10% to ±20% | Non-linear | Cost-sensitive SMPS, non-critical power rails |
| Powdered Iron | Iron powder suspended in phenolic binder, usually toroidal | ±10% to ±15% | Highly stable up to 100°C | Power Factor Correction (PFC), high-current output chokes |
| Ceramic Core (LTCC) | Multilayer Low-Temperature Co-fired Ceramic or alumina wirewound | ±2% to ±10% | +100 to +250 | GHz RF signal routing, impedance matching networks |
According to the Coilcraft Learning Center, the most common mistake designers make is selecting an unshielded drum core for a mixed-signal board. The fringing magnetic flux from an unshielded inductor can easily induce noise into nearby high-impedance analog traces or Hall-effect sensors. Always pay the 10-15% cost premium for shielded ferrite when routing near ADCs.
Decoding the Markings: What the Dots and Numbers Actually Mean
Unlike resistors, where the 3-digit code is universally understood in ohms, SMD inductor markings trip up even experienced engineers because the base unit is almost always microhenries (µH), and the multiplier logic has exceptions.
The 3-Digit SMD Code
- 100: 10 × 10⁰ = 10 µH (Not 100 µH!)
- 101: 10 × 10¹ = 100 µH
- 472: 47 × 10² = 4700 µH (or 4.7 mH)
The 'R' Decimal Indicator
For values under 10 µH, the letter 'R' replaces the decimal point.
- 4R7: 4.7 µH
- R47: 0.47 µH
- 1R0: 1.0 µH
The Dot Convention (Polarity)
If you see a white or copper-colored dot on one end of the inductor, it indicates Pin 1. For standard power inductors, polarity rarely matters. However, for coupled inductors (used in SEPIC or Zeta converters) and common-mode chokes, the dot defines the phase relationship. Reversing the phase on a coupled inductor in a SEPIC topology will result in the switching node ringing violently and destroying the MOSFET. Always verify the dot against the datasheet's schematic symbol.
Bench Autopsy: When Inductors Fail and How to Spot It
Inductors rarely fail silently. When they do, the physical evidence on the PCB tells you exactly what went wrong.
| Failure Mode | Visual Symptom | Electrical Measurement | Root Cause |
|---|---|---|---|
| Thermal Overload (Open) | Darkened, blistered epoxy; cracked shield can; smell of burning phenolic. | OL (Open Loop) on DCR test. | RMS current exceeded wire ampacity; enamel insulation melted and fused, then burned open. |
| Mechanical Fracture | Hairline crack in the ferrite drum or toroid, often near the PCB pad. | Inductance drops by 20-50%. | Board flexure during depaneling or drop shock. The crack introduces an unintended air gap. |
| Core Saturation | No visual damage to the inductor. The switching IC next to it is cratered or split. | Inductor tests fine on an LCR meter at 1kHz. | Peak current exceeded Isat. Inductance collapsed to near-zero, causing a massive di/dt spike that punched through the IC's silicon. |
War Story: The 100kHz Buck Converter Meltdown
Let us look at a real-world scenario that highlights the difference between thermal limits and magnetic limits. I was prototyping a 12V-to-5V, 3A buck converter using the classic LM2596 (150kHz switching frequency).
The Setup: Using the standard inductor sizing formula, I calculated a required inductance of roughly 22 µH. To keep ripple low, I bumped up to a standard 33 µH value. I pulled an unshielded SMD drum inductor from the bench bin labeled "330" (33 µH) with an Irms (thermal rating) of 3.5A. Since my max load was 3A, I figured I had a comfortable 0.5A margin.
The Numbers: The LM2596 has a peak current limit of roughly 4.5A. In a buck converter, the peak inductor current is $I_{load} + \frac{\Delta I_L}{2}$. At 3A load, my peak current was hitting about 3.6A.
The Outcome: The moment I drew 2.8A from the output, the LM2596 popped with an audible crack, and the chip package split. The inductor looked perfectly fine.
What Went Wrong: I had confused $I_{rms}$ (the thermal limit of the copper wire) with $I_{sat}$ (the magnetic saturation limit of the ferrite core). The bin inductor I grabbed had an $I_{sat}$ of only 2.5A. When the peak current hit 2.5A, the ferrite core saturated. The inductance instantly dropped from 33 µH to essentially the parasitic inductance of a straight wire (a few nanohenries). The current ramped up at an uncontrollable $di/dt$, bypassing the LM2596's thermal shutdown (which is too slow to catch nanosecond spikes) and melting the internal power MOSFET. Always check $I_{sat}$, not just $I_{rms}$.
The Substitution Matrix: Swapping Parts Without Frying the Board
Supply chain shortages mean you will frequently need to substitute inductors. According to Texas Instruments power design guidelines, magnetics are the most sensitive components to substitute blindly. Follow this numbered protocol to ensure your swap will not compromise reliability or EMI compliance.
- Match Inductance (±20% for Power, ±5% for RF): In a DC-DC converter, a 20% deviation in inductance will simply shift your ripple current and slightly alter the crossover frequency of the control loop. It is usually survivable. In an RF matching network, a 10% shift will detune your VSWR and kill your transmit power.
- Verify $I_{sat}$ > $I_{peak}$ (Non-Negotiable): The substitute's saturation current must exceed the peak current of your circuit, including transient load steps and the IC's internal current limit threshold. If the original was 4A Isat, a 3.5A Isat substitute is a ticking time bomb.
- Check DCR and Control Loop Stability: Lower DCR is generally better for efficiency. However, if you are using a peak-current-mode controller that relies on the inductor's DCR for current sensing (DCR sensing), substituting an inductor with significantly lower DCR will reduce the sense voltage below the controller's noise floor, causing sub-harmonic oscillation.
- Evaluate Shielding Requirements: Never substitute an unshielded inductor for a shielded one if the component is within 10mm of a sensitive analog trace, a magnetic sensor, or an unshielded enclosure seam. The All About Circuits AC textbook notes that fringing flux follows the inverse-cube law; moving it 5mm away might not be enough to save your noise floor.
- Footprint and Pad Geometry: Even if the electrical specs match, a substitute with a different physical pad layout can cause tombstoning during reflow or create a mechanical weak point under vibration. Check the 3D STEP model before approving the BOM swap.
Inductors are not passive afterthoughts; they are the magnetic heart of your power and signal chains. Respect the core material, verify the saturation limits, and read the dots. Your silicon will thank you.
