Electromagnets vs. Permanent Magnets
One makes its field from electric current, the other from the material itself — and that single difference drives every trade-off: control, power, heat, reliability, and cost. Here's how to choose, plus the electro-permanent hybrids that split the difference.
Permanent magnet — when the field should always be on: holding, sensing, coupling, motor rotors, separation. Zero power, zero heat, nothing to fail, smallest package for the field.
Electromagnet — when you must switch, vary, or reverse the field: lifting and release, actuation with proportional control, AC devices, very large working volumes.
Electro-permanent hybrid — when you need switchable holding without continuous power: the magnet holds, a brief pulse toggles it. Increasingly the right answer for workholding, lifting, and end-of-arm tooling.
The fundamental difference
An electromagnet generates its field from current flowing in a coil, almost always wrapped around a soft-iron core that multiplies and guides the flux. Current on → field on; current off → field gone (minus a little residual). Field strength follows current, so it is adjustable, reversible, and controllable in real time — at the price of continuous power and the resistive heat that comes with it.
A permanent magnet stores its field in the material's magnetic structure — set once during manufacture (see how NdFeB magnets are made) and maintained indefinitely with no energy input. The field is always on, essentially fixed, and can only be removed by abusing the magnet: overheating it, opposing fields beyond its coercivity, or physical damage. Modern NdFeB properly specified loses only a small fraction of its flux over decades.
An electromagnet is a field you pay for by the second. A permanent magnet is a field you paid for once, in the material. Every row of the comparison table below is that sentence wearing different clothes.
Head-to-head comparison
| Attribute | Electromagnet | Permanent magnet (NdFeB) |
|---|---|---|
| Field control | Fully adjustable, reversible, switchable | Fixed (mechanical shunting/rotation for crude switching) |
| Power consumption | Continuous while energized | Zero, forever |
| Heat generation | I²R losses; duty-cycle and cooling limited | None |
| Field strength (compact scale) | Modest without large coils/cooling | Very high per unit volume — up to ~1.4 T at the pole of good circuits |
| Field strength (no size limit) | Unlimited in principle (superconducting: many tesla) | Practically capped (Halbach cylinders reach a few tesla) |
| Response to power loss | Field drops out — load releases | Nothing changes — inherently fail-safe holding |
| Size & weight for a given field | Larger, heavier (copper + iron + supply) | Smallest and lightest option |
| Supporting equipment | Power supply, wiring, control, thermal management | None |
| Wear-out / failure modes | Insulation aging, coil burnout, supply failure | None in normal service; loss only via heat/field/damage abuse |
| Temperature ceiling | Insulation class limited (~155–180 °C typical) | Grade dependent, 80–230 °C for NdFeB, 350 °C+ for SmCo — see grades chart |
| Unit cost | Coil + core + supply + controls | Just the magnet (see pricing guide) |
Where electromagnets win
- Anything that must release on command: scrapyard lifting magnets, door holders that release on fire alarm, brakes and clutches, sorting and diverting.
- Proportional control: solenoid valves and actuators, active magnetic bearings, focusing and steering coils, voice-coil positioning — wherever force must track a signal.
- Alternating or rotating fields by design: transformers, induction heating, motor stators — the AC half of every electric machine is necessarily an electromagnet.
- Very large or very strong working fields: MRI main fields (superconducting), research magnets, big separators — where the coil scales and the material can't.
- Degaussing and magnetizing: creating and erasing magnetization is inherently an electromagnet's job — including the fixtures that charge permanent magnets.
Where permanent magnets win
- Always-on functions: holding, latching, closures, mounting, magnetic separation, couplings and gears — paying continuous watts for a constant field is waste.
- Sensing: position, speed, and angle sensors use a PM field precisely because it is stable, compact, and unpowered — see magnetization directions for the sensor configurations.
- Motor and generator rotors: the PM rotor (paired with an electromagnet stator) is the dominant architecture for efficiency-critical machines from drones to EVs — the field that never needs slip rings or excitation current.
- Battery, remote, and sealed devices: zero power and zero maintenance make PMs the only option where energy or access is scarce.
- Fail-safe holding: when power loss must NOT drop the load — safety interlocks, fail-secure latches — the permanent magnet holds and energy is applied only to release (the permanent-magnet brake pattern).
- Weight- and space-critical designs: per newton of force or gauss at the gap, a modern NdFeB circuit is dramatically smaller and lighter than a coil — see the pull force guide for what those forces look like.
Electro-permanent hybrids
The most interesting territory is in between — devices that hold like a permanent magnet and switch like an electromagnet:
- Electro-permanent magnets (EPMs): pair a high-coercivity magnet (NdFeB) with a low-coercivity switchable magnet (AlNiCo) in one circuit. A brief coil pulse flips the AlNiCo's polarity, steering the combined flux either into the workpiece (ON) or into an internal loop (OFF). Power is consumed only during the millisecond switch — holding is free and fail-safe. Now standard in machine-tool workholding, mold clamping, robot end-of-arm tooling, and drone cargo latches.
- Permanent-magnet-biased devices: PM provides the standing force, a small coil adds or subtracts — PM brakes (spring-set alternatives), latching solenoids and relays that hold position unpowered, magnetic bearings with PM bias to slash coil current.
- Mechanically switched PM bases: the classic on/off magnetic base rotates one magnet against another to cancel or combine flux — no electricity at all, human-powered switching.
If your electromagnet spends most of its life energized at constant current just to hold something, it's a permanent magnet or EPM application wearing the wrong hardware — and the energy, heat, and fail-drop behavior are all fixable in one redesign.
Where they work together
The framing "versus" undersells how often the answer is both: nearly every modern electric machine is a permanent magnet and an electromagnet in partnership. The BLDC/PMSM motor is a PM rotor driven by electromagnet stator coils; a voice coil actuator is a coil moving in a PM field; a loudspeaker is the same device making sound. In each, the PM supplies the constant field for free and the coil supplies the controlled part — each technology doing exactly what it's best at.
For rotor-side options in those machines — segmented, true radial rings, and Halbach arrays — the field-shaping choices are a permanent-magnet design discipline of their own.
Cost & energy over a product's life
Unit-price comparisons flatter electromagnets less than they appear to once the system bill arrives:
- Electromagnet system cost = coil + core + power supply + control + wiring + thermal management + the enclosure space for all of it — then operating energy for every hour energized, for the life of the product.
- Permanent magnet system cost = the magnet (and perhaps a steel return path), once. A holding magnet drawing zero watts for ten years against a 20 W holding coil is a four-figure energy difference per unit at industrial duty cycles.
- Where the electromagnet earns it back: when controllability replaces mechanisms — one coil replacing an actuator, latch, and sensor can win the system-cost comparison decisively. The honest accounting is always at system level.