Transformer Lamination Theory: Why Core Thickness Dictates Heat and Hum

Transformer laminations are thin, insulated sheets of silicon steel stacked together to form a magnetic core, designed to minimize energy-wasting eddy currents. In a real circuit or installation, the lamination stack dictates your core losses (efficiency), operating temperature, physical weight, and acoustic hum. Makers and junior engineers commonly confuse lamination thickness with core material—assuming all steel cores behave identically at 60Hz—or they blame hysteresis losses when eddy currents are actually the culprit cooking their windings.

The Physics of the Stack: Breaking the Eddy Current Path

Faraday’s law of induction doesn't just apply to your copper windings; it applies to the iron core sitting inside them. When alternating magnetic flux passes through a conductive core, it induces a voltage within the core material itself. If the core is a solid block of iron, this induced voltage drives massive, swirling electrical currents—called eddy currents—through the metal. Because iron has low electrical resistance, these currents generate immense heat via I²R losses, potentially melting the transformer or tripping your bench breaker.

To stop this, we slice the core. Imagine water flowing down a wide, smooth concrete spillway versus a spillway broken up by a series of vertical baffles. The baffles don't stop the water from falling (representing the magnetic flux passing through the core), but they break up the massive, destructive swirling eddies (the electrical currents) into tiny, manageable ripples.

By stacking thin sheets of steel coated in an insulating varnish or oxide layer, we force the eddy currents to remain confined within the tiny cross-sectional area of each individual sheet. Since electrical resistance is inversely proportional to cross-sectional area, the thinner the lamination, the higher the resistance to eddy currents, and the lower the heat generated.

The Math: Lamination Thickness and Core Loss

The relationship between lamination thickness and heat is not linear; it is exponential. The power lost to eddy currents ($P_e$) is proportional to the square of the lamination thickness ($t$). The governing proportionality is:

$P_e \propto K_e \cdot f^2 \cdot B_{max}^2 \cdot t^2$

Where $f$ is frequency, $B_{max}$ is maximum flux density, and $K_e$ is a material constant. Because thickness is squared, halving your lamination thickness doesn't halve your eddy current loss—it quarters it.

Where You Meet Transformer Laminations in Practice

You will encounter lamination stacks primarily in low-frequency (mains and audio) applications. The specific thickness and alloy you need depend entirely on your operating frequency.

• Mains Frequency (50/60Hz): You will use silicon steel laminations ranging from 0.23mm to 0.35mm thick. These are found in heavy EI, UI, and toroidal transformers for linear power supplies, HVAC contactors, and grid distribution.
• Audio Frequency (20Hz–20kHz): Standard silicon steel rolls off too early and introduces phase shift. Audio output transformers use nickel-iron alloys (like Mu-metal or Permalloy) with ultra-thin laminations (0.1mm or less) to maintain high permeability at low flux densities without high-frequency attenuation.
• High Frequency (>20kHz, SMPS): Laminations are abandoned entirely. At 100kHz, even 0.1mm steel would melt from eddy currents. Switch-mode power supplies use powdered iron or ceramic ferrite cores, which are inherently non-conductive and immune to bulk eddy currents.

According to core loss models detailed by All About Circuits, failing to match the core material to the frequency guarantees catastrophic thermal runaway. You cannot use a 60Hz laminated iron core in a 50kHz inverter circuit.

Bench Scenario: The Buzzing, Overheating DIY Linear PSU

Theory is clean; the workbench is not. Here is a classic failure mode that ruins weekends and melts solder.

The Setup

A builder is constructing a 50V, 10A linear power supply for a Class AB audio amplifier. To save money, they salvage a heavy EI core from a discarded microwave oven transformer (MOT) and wind new secondaries. To "secure" the heavy core and prevent buzzing, they wrap the outside of the lamination stack tightly in bare aluminum flashing and bolt it down to the chassis using steel L-brackets that press directly against the core edges.

The Numbers

MOT cores are designed for high flux density but very low duty cycles (a microwave runs for minutes, not hours). The builder pushes the core to 1.6 Tesla continuously. The primary winding is rated for 120V AC.

The Outcome

Within 15 minutes of powering on, the transformer is screaming with a violent 120Hz acoustic hum. The aluminum wrap is too hot to touch, and the primary draws 4.5 Amps at no-load (a healthy 1kVA transformer should draw less than 0.2A at no-load). The varnish on the copper windings begins to blister.

What Went Wrong

The builder accidentally created a shorted turn across the outside of the lamination stack. While the internal factory varnish was intact, the bare aluminum wrap and steel brackets created a highly conductive path bridging the edges of the laminations. This allowed massive eddy currents to flow laterally across the entire stack boundary, effectively turning the outer shell of the core back into a solid block of metal. Furthermore, the steel brackets provided a secondary path for flux to leak and induce currents into the chassis.

The Fix: Never use conductive materials to wrap or clamp a lamination stack. Remove the brackets, use nylon or plastic bobbin end-frames, and clamp the EI stack using non-magnetic, non-conductive tension bands or specialized steel brackets that are physically separated from the core edges by insulating Nomex paper.

Core Material and Lamination Reference Chart

When sourcing replacement laminations or specifying a custom transformer, use this reference chart to match the material to your application. Data aligns with standard Electronics Tutorials core loss specifications.

FAQ: Transformer Lamination Gotchas