There are exactly five fundamental types of magnetism—diamagnetic, paramagnetic, ferromagnetic, antiferromagnetic, and ferrimagnetic—but only three directly dictate everyday electrical and electronic component selection. Magnetism is the physical phenomenon arising from electron spin and orbital motion that determines how a material reacts when exposed to an external magnetic field. In a real circuit or installation, the specific type of magnetism a material exhibits dictates your core losses (eddy currents), inductance density, thermal limits, and high-frequency shielding effectiveness. The most costly mistake hobbyists and junior engineers make is confusing ferromagnetic materials (like solid iron or silicon steel) with ferrimagnetic materials (like ceramic ferrites), failing to realize that ferrites are electrically insulating, which is the only reason they don't melt down from eddy currents in high-frequency switching power supplies.

The 5 Types of Magnetism: A Spec-Sheet Breakdown

Before you can select a core for an inductor or a shielding material for a sensitive analog front-end, you need to know how these five categories behave at the atomic level and how they translate to bench-level components. Below is the definitive reference table for magnetic material types.

Magnetic Type Relative Permeability ($\mu_r$) Response to External Field Common Electrical Materials Primary Circuit Application
Diamagnetic Slightly less than 1 (e.g., 0.99999) Weakly repelled; creates an opposing internal field. Copper, Bismuth, Gold, Superconductors RF trace routing, superconducting MRI magnets, Meissner effect levitation.
Paramagnetic Slightly greater than 1 (e.g., 1.00002) Weakly attracted; thermal agitation disrupts alignment. Aluminum, Platinum, Liquid Oxygen Generally avoided in magnetics; used in specialized cryogenic sensors.
Ferromagnetic High (1,000 to 400,000+) Strongly attracted; retains magnetization (hysteresis). Iron, Cobalt, Nickel, M19 Silicon Steel, Mu-metal 50/60Hz mains transformers, motor stators, low-frequency EMI shielding.
Antiferromagnetic Very low (close to 1) Adjacent atomic moments align anti-parallel, canceling out. Chromium, Manganese Oxide (MnO) Exchange bias layers in spin-valve read heads (hard drives, MRAM).
Ferrimagnetic Moderate to High (20 to 5,000) Anti-parallel alignment but unequal moments yield net magnetization. Magnetite, MnZn Ferrite, NiZn Ferrite, YIG Switch-mode power supplies (SMPS), RF chokes, EMI suppression beads.
Bench Note: Notice the jump in resistivity between Ferromagnetic and Ferrimagnetic materials. While both have high permeability, bulk iron has a resistivity of roughly $10^{-7} \Omega\cdot m$, whereas MnZn ferrite sits around $10 \Omega\cdot m$. That eight-order-of-magnitude difference in electrical resistance is what allows ferrites to operate at 500 kHz without turning into a toaster.

Worked Example: Core Loss and Saturation at 100 kHz

Let’s look at a real-world scenario: you are designing a 100 kHz buck converter and need to choose an inductor core. You have two options on the bench: a scrap piece of M19 Grain-Oriented Silicon Steel (ferromagnetic) from a microwave transformer, and a TDK N87 Toroid (ferrimagnetic ferrite).

Both cores have the same physical dimensions and an effective cross-sectional area ($A_e$) of $1.0 \text{ cm}^2$. You need to pass a peak flux density ($B_{pk}$) of 0.2 Tesla.

1. The Eddy Current Loss Calculation:
Eddy current power loss ($P_e$) in a core is proportional to the square of the frequency ($f$), the square of the lamination thickness ($t$), and inversely proportional to the material resistivity ($\rho$). The simplified proportionality is:

$P_e \propto \frac{t^2 \cdot f^2 \cdot B_{pk}^2}{\rho}$
  • M19 Silicon Steel: Standard lamination thickness $t = 0.35 \text{ mm}$ ($3.5 \times 10^{-4} \text{ m}$). Resistivity $\rho \approx 4.7 \times 10^{-7} \Omega\cdot m$.
  • N87 Ferrite: Because it is a ceramic, eddy currents are confined to microscopic grain boundaries. We treat the bulk resistivity as $\rho \approx 10 \Omega\cdot m$, and the effective "thickness" is the grain size (negligible).

If you run the M19 steel at 100 kHz, the $f^2$ term ($10^{10}$) combined with the tiny resistivity ($10^{-7}$) results in massive eddy currents. The core will dissipate hundreds of watts per kilogram, rapidly exceeding the Curie temperature ($\sim 740^\circ\text{C}$ for steel) and destroying the winding insulation. The N87 ferrite, with a resistivity $10^7$ times higher, will only dissipate roughly $50 \text{ mW/cm}^3$ at the same flux density and frequency—easily managed by ambient air cooling.

2. The Saturation Trap:
Here is where the ferrimagnetic material bites back. M19 steel saturates at roughly 1.8 Tesla. N87 ferrite saturates at just 0.39 Tesla at room temperature, and that drops to 0.25 Tesla at 100°C. If you attempt to push 0.5 Tesla through the N87 core to shrink your inductor size, the permeability will cliff-drop to near 1, your inductance will vanish, and your switching MOSFETs will instantly short-circuit and explode due to unconstrained $di/dt$.

Where You Meet This in Practice

Understanding these magnetic types isn't just academic; it dictates the bill of materials for almost every power and signal system you will build.

Mains Wiring and 60Hz Transformers (Ferromagnetic)

When you wire a 240V to 120V step-down transformer or spec a subpanel feeder, the core inside is almost certainly grain-oriented electrical steel (GOES). Ferromagnetic materials are mandatory here because the 60Hz frequency is low enough that eddy currents can be managed simply by slicing the core into thin, varnished laminations. You need the massive 1.5T+ saturation limit of iron to handle the high volt-seconds of a 50/60Hz sine wave without requiring a core the size of a car engine.

Switch-Mode Power Supplies and RF (Ferrimagnetic)

Open up a modern laptop charger or a 2026-era GaN USB-C fast charger, and you will find ferrimagnetic cores. Specifically, MnZn (Manganese-Zinc) ferrites are used for power conversion up to 2 MHz due to their high permeability ($\mu_i \approx 2000$), while NiZn (Nickel-Zinc) ferrites ($\mu_i \approx 100-800$) are used for EMI suppression beads and VHF/UHF chokes because their resistivity is even higher, preventing parasitic capacitance losses at extreme frequencies.

Precision Analog Shielding (Diamagnetic & High-Permeability Ferro)

If you are building a low-noise preamplifier for a load cell or an EEG sensor, you must shield against 60Hz magnetic interference. Standard aluminum enclosures (paramagnetic) and copper foil (diamagnetic) do absolutely nothing to block low-frequency magnetic fields; they only block electric fields and high-frequency RF via the Faraday cage effect. To block 60Hz magnetic flux, you must use a high-permeability ferromagnetic alloy like Mu-metal ($\mu_r > 80,000$), which acts as a low-reluctance shunt, routing the magnetic field lines around your sensitive traces rather than through them.

Fatal Design Mistakes: Swapping Ferro for Ferri

When repairing or modifying magnetics, builders frequently make substitutions that lead to catastrophic failure. Keep these rules on your bench:

  1. Never use a ferrite core for a 60Hz inverter output filter. The low frequency requires a high volt-second capacity. The ferrite will saturate in the first quarter-cycle of the 60Hz wave, turning your inductor into a dead short and blowing your H-bridge.
  2. Never use solid iron or steel bolts as inductor cores for RF circuits. The solid ferromagnetic mass will generate immense eddy currents, acting as a shorted secondary turn. It will heat up to glowing red and melt your solder joints within seconds.
  3. Beware of the Curie Temperature ($T_c$). Ferrimagnetic materials lose all magnetic properties above their Curie temperature (often just $100^\circ\text{C}$ to $200^\circ\text{C}$ for power ferrites). If your SMPS transformer runs too hot, it doesn't just degrade; it completely loses its inductance, resulting in an immediate, hard-failure short circuit.

Frequently Asked Questions

Can I use a neodymium magnet as a transformer core?
No. Neodymium (NdFeB) is a permanent magnet (hard ferromagnetic). It has a fixed, massive internal magnetic field and very low incremental permeability. A transformer core requires a "soft" magnetic material that can easily reverse its magnetic domains with alternating current. Using a permanent magnet will result in immediate saturation and zero energy transfer.

Why do ferrite beads get hot on PCB power rails?
Ferrite beads are designed to be lossy at high frequencies, converting RF noise into heat. However, if you pass too much DC current through them, the internal magnetic field can partially saturate the ferrimagnetic material, dropping its impedance and causing excessive $I^2R$ heating from the DC resistance of the winding.

For deeper material specifications, always consult the manufacturer's databooks, such as the TDK Ferrites catalog for exact $B-H$ curves and loss tangents, or reference the foundational physics models at Georgia State University's HyperPhysics. When designing magnetics, the material type isn't just a suggestion; it is the absolute physical boundary of your circuit's performance.