Inside the Flyback Transformer: The Component That Defines a Small Driver

Open an LED driver and the largest component inside it is almost always the transformer. In a plastic-housed 100W unit, the transformer is the block of ferrite and copper that carries the isolation between the mains side and the SELV output, stores energy, and sets the operating point of the whole converter. It is also the component whose design determines whether the driver is efficient, quiet, compact, and safe. The flyback topology used in most constant voltage drivers (introduced earlier as one of the basic circuit families) is defined by this single component, and understanding what goes into its design explains many of the specifications on a datasheet.

The flyback transformer works differently from a conventional power transformer. In a normal AC transformer, power transfers continuously from primary to secondary through the core. In a flyback, the primary switch stores energy in the core's magnetic field when it is on, and that energy is released to the secondary when the switch turns off. The transformer is really an inductor with a controlled air gap and a second winding. The air gap is essential: without it, the ferrite saturates quickly and cannot store the required energy. The gap sets the inductance of the primary winding and controls how the stored energy transfers to the output.

The primary inductance is the first design choice. It determines how much energy the transformer stores per switching cycle: E = ½ × Lp × Ipeak². A larger Lp means lower peak current and lower conduction loss, but a larger, more expensive core. A smaller Lp means higher peak current, more copper loss, and more noise. The switching frequency (typically 65kHz to 140kHz for a small driver) sets how much energy must be stored per cycle: at higher frequency, each cycle stores less energy, so the transformer can be smaller — the same power in a smaller core. This is the core trade-off that makes a 100W driver the size of a paperback book rather than the size of an old linear-supply transformer.

The turns ratio sets the output voltage. The input bus (around 300 to 400V DC after rectification) and the desired output (24V) define the ratio: Np/Ns must satisfy the volt-second balance of the core, accounting for the diode drop, the switch voltage, and the transformer losses. A higher ratio means fewer secondary turns, which is good for copper but raises the secondary peak current. The transformer designer picks the ratio to balance primary and secondary losses and to keep the voltage stress on the output diode and the primary switch within safe limits. This is why a 24V driver and a 12V driver of the same power rating often use different transformer windings, not a different secondary regulator.

Flux density is the safety margin that prevents saturation. The core's flux density swings with each switching cycle; if it swings too far, the ferrite saturates, the primary current spikes out of control, and the switch fails. A typical ferrite core runs at 100 to 200 millitesla peak, well below its saturation flux density of around 300 to 400mT at room temperature — but saturation drops as the core heats, which is why the margin is sized at the worst-case temperature and load. The audible noise discussed earlier is also a transformer phenomenon: the magnetostriction of the ferrite at the switching frequency, and the mechanical coupling of the windings to the bobbin. A well-designed transformer has low flux density, tight winding construction, and a core that does not vibrate audibly.

For a buyer, the transformer matters indirectly. A driver built around a properly sized transformer — adequate core, conservative flux density, good winding isolation — runs cooler, is quieter, and lasts longer. A driver that squeezes the same power into a smaller transformer to save cost trades efficiency, thermal margin, and audible noise for a smaller plastic case. The transformer is where the efficiency and the Tc rating on the nameplate actually come from. Asking about the core size, the switching frequency, and the isolation winding construction separates a driver built on a proper magnetic design from one that merely meets the power number on paper.