Two ratings, two different failure mechanisms
An inductor datasheet usually lists two current figures, and reading one as though it were the other is one of the more expensive simplification errors in power design. They describe genuinely different limits:
- Saturation current (Isat) — the DC current at which inductance has fallen to a defined fraction of nominal, because the core has begun to saturate. This is a magnetic limit.
- RMS or temperature-rise current (Irms) — the DC current that raises the component's temperature by a specified amount, usually quoted as a rise above ambient, caused by copper loss in the winding. This is a thermal limit.
Exceed the saturation current and the inductance collapses — in a switching converter the current can then rise much faster than the control loop expects. Exceed the RMS current and the part overheats, which degrades insulation, damages the winding, and shifts the resistance further. One is an electrical failure, the other a thermal one, and the datasheet may give no hint as to which is lower for the part you are looking at.
Peak current is not load current
In a switching regulator the current in the inductor is not a steady value. It ramps up while the switch is on and down while it is off, so it sits at the output current on average but reaches a peak above it:
peak current = the DC current, plus half the peak-to-peak ripple
Because the ripple depends on inductance, input and output voltage and switching frequency, the peak can be substantially above the average. Comparing Isat against the load current alone therefore leaves the ripple and the tolerance on it unaccounted for — a margin that looks generous on paper can be negative in the circuit.
There is a second transient worth considering. A start-up inrush into a large output capacitance, or a short circuit before protection reacts, can produce a current spike well above the steady-state peak. Where the protection is not fast, the inductor may see it first.
Where the thermal limit bites
The RMS rating is usually quoted for a specific temperature rise in a specific mounting condition, and the mounting condition is easy to overlook. An inductor sitting in still air inside a closed enclosure is not in the condition the datasheet assumed, and nearby heat sources — a power switch, a diode, a transformer — raise the starting temperature before the part's own losses are added.
Copper resistance also rises with temperature, so a hot inductor is a lossier inductor, whose extra loss makes it hotter. The effect is bounded by the physics but it is not in your favour, and it is why the placement and the copper area under the part matter as much as the part number.
DC resistance and efficiency
DC resistance (DCR) is the winding's resistance, and it turns directly into conduction loss proportional to the square of the current. In a high-current rail that loss can dominate the whole converter's efficiency, and it appears as heat inside the smallest component in the circuit. Where efficiency matters, DCR is often the parameter to optimise first, before inductance tolerance.
Shielded or unshielded
A shielded inductor contains its magnetic field. An unshielded one radiates, and the field can couple into nearby traces, into sensitive analogue nodes, and into other magnetics. Three practical consequences:
- EMI. Unshielded parts can contribute to radiated emissions and make compliance harder, often at the frequencies you were not looking at.
- Placement becomes a constraint. An unshielded part next to a feedback node or a sensor can inject noise. A shielded part relaxes that constraint at the cost of size and price.
- Mutual coupling. Two unshielded inductors close together can interact, changing the effective inductance of both — which is how a layout change alters behaviour nobody expected to change.
Self-resonant frequency
Every winding has inter-turn capacitance, so every inductor has a self-resonant frequency. Below it the part behaves inductively; above it the capacitance dominates and the part behaves capacitively, so it no longer attenuates and may even amplify noise. The switching frequency and its harmonics need to stay comfortably below that point, which is another reason a physically large, high-inductance part is not automatically the better choice.
Inductance tolerance, and the design margin it implies
Inductance tolerance on power parts is commonly wide — tens of percent is normal. Design calculations should tolerate the range, not the nominal value, and the ripple current moves with inductance in the opposite direction, which interacts with the peak-current question above.
A ferrite bead is not an inductor
The two are often confused because both are magnetic components and both appear near power rails. They are built for opposite jobs:
- An inductor stores energy, and is specified by inductance and by current ratings.
- A ferrite bead is deliberately lossy, converting high-frequency noise into heat. It is specified by impedance at a given frequency, and it is not intended to store energy or carry a large DC current without derating.
Using a bead where an inductor belongs, or the reverse, is a category error that no amount of value tuning will fix.
Common selection errors
- Comparing saturation current against the average load current, not the peak
- Assuming the saturation and RMS ratings are the same figure under different headings
- Ignoring the mounting condition and nearby heat when reading the temperature-rise rating
- Choosing an unshielded part for a noise-sensitive location
- Operating near self-resonance, or at a switching harmonic above it
- Designing around nominal inductance and ignoring a wide tolerance
- Substituting a ferrite bead for an inductor, or the reverse
Ordering magnetics
At YILUDA TECH every listed part is 100% original, new and unused, with no refurbished or counterfeit material, held in stock at our Shenzhen warehouse and shipped worldwide by DHL, FedEx or UPS. Incoming inspection and traceability documentation are available on request.
Send the full ordering code with quantity and destination and we will confirm availability, pricing and lead time. Because current rating, shielding and tolerance all sit in the suffix, our guide to reading a part number is worth a look first. Browse the full component inventory, or start from Coilcraft, Murata or Vishay.