Ferrite Cores for Isolated DC-DC Converters

Isolation is a safety and system requirement in a large share of power supplies, and it is provided by the transformer rather than by the switching devices. That makes the transformer the component that defines the converter as much as the control scheme does — and the topology chosen determines what the core has to do.

The most useful way to think about isolated topologies is whether the transformer stores energy or transfers it. Gapped flyback and forward converters store energy and must be designed to handle a DC flux component, while bridge and push-pull converters transfer energy and operate with symmetrical flux excursions.

Flyback converters

The flyback converter stores energy in the transformer during the switch-on interval and delivers it to the secondary during switch-off. The transformer therefore functions as a coupled inductor rather than as a true transformer, and the core must store the full energy transferred each cycle.

That energy storage requirement means the core is gapped. The gap reduces effective permeability so the core can store the required energy without saturating, and the design must account for the DC bias that results from unidirectional current. Flyback suits low to moderate power and is common in auxiliary supplies and chargers.

Forward converters

The forward converter transfers energy during the switch-on interval through a true transformer action, with an output inductor providing the energy storage for the load. The transformer core still needs a reset mechanism to remove the magnetizing flux each cycle, which is commonly provided by a reset winding or by active clamping.

Because the core resets each cycle, it does not need to store the transferred energy, but the reset interval limits the maximum duty cycle. The design has to keep the flux excursion within the available range so the core returns to its starting point before the next cycle begins.

Push-pull, half-bridge and full-bridge

These topologies drive the transformer alternately in both directions, so the flux swings symmetrically around zero and the core operates as a true transformer. That allows higher permeability to be used and generally produces smaller magnetics for a given power level.

• Symmetrical drive means no DC flux component in steady state
• Core is typically ungapped, since energy is transferred not stored
• Saturation margin is required for imbalance from component tolerance
• Half bridge and full bridge suit higher power levels
• Push-pull suits lower input voltages, with attention to switch voltage stress

Even with symmetrical drive, practical designs keep margin below the saturation limit. Slight differences in switch timing, winding resistance or device characteristics can produce a small flux offset, and the core must tolerate that without saturating.

Why ferrite dominates isolated converters

Ferrite combines high permeability with low core loss at the frequencies isolated converters typically use, and its high resistivity limits eddy-current loss. Manganese-zinc grades are the conventional choice for power conversion in the tens to hundreds of kilohertz; nickel-zinc grades become relevant at higher frequencies and lower power levels.

Ferrite also offers a wide range of core shapes, which matters when the transformer must fit a specific footprint and satisfy isolation distances at the same time.

Engineering requirements to define

• Topology and whether the core stores or transfers energy
• Switching frequency and duty cycle range
• Input and output voltage and power level
• Whether the core requires a gap and what inductance that implies
• Peak flux density with margin below saturation
• Core loss at the operating frequency and flux swing
• Isolation and creepage distances required
• Temperature rise and cooling method

Design considerations

Isolation requirements frequently drive the physical design more than the electrical ones do. Creepage and clearance distances determine how far the windings must be separated, which affects the winding window, the leakage inductance and the achievable coupling. A core shape chosen purely for electrical performance may not accommodate the isolation barrier.

Thermal behavior follows from loss, and loss comes from both the core and the windings. At higher frequencies the proximity effect in the copper can dominate, which is why litz wire and careful layer arrangement are common in compact isolated converters.

Application examples

• Auxiliary and bias supplies requiring isolation
• Industrial DC-DC modules with wide input range
• Telecom and server intermediate bus converters
• Battery chargers and on-board charger stages
• Medical and instrumentation supplies with stringent isolation requirements

Selection checklist

• Identify whether the topology stores energy, which decides gapped or ungapped
• Size the core with margin below saturation for imbalance and transients
• Confirm the reset mechanism for single-ended topologies
• Evaluate core loss at the real operating point
• Check the core shape accommodates the isolation barrier
• Verify temperature rise at worst-case ambient and load

Custom development for isolated converters

Isolated converters are usually built around a transformer developed for the specific voltage, power and isolation requirements, because catalogue parts rarely match all three. Development covers material grade, core shape, gap where required, winding arrangement and isolation design, followed by sampling and qualification.

See Ferrite Cores for DC-DC Converters for the non-isolated case, Ferrite Cores for Phase-Shifted Full-Bridge Converters for a soft-switched bridge topology, and the Ferrite Core Guide for materials and shapes.

Frequently Asked Questions

Why do isolated DC-DC converters need a transformer?

The transformer provides galvanic isolation between input and output, which is a safety requirement in many applications, and it also performs voltage transformation. The isolation is provided by the magnetic coupling rather than by the switching devices.

Which isolated topologies need a gapped core?

Single-ended topologies that store energy in the transformer, such as the flyback and the forward converter, generally need a gapped core. Bridge and push-pull topologies transfer energy with symmetrical flux excursions and typically use an ungapped core.

Why is a flyback transformer described as a coupled inductor?

Because it stores energy during the switch-on interval and releases it to the secondary during switch-off, rather than transferring energy continuously as a true transformer does. That is why its core must be gapped to store the required energy without saturating.

Is flux imbalance a concern in bridge converters?

It is in practice. Symmetrical drive means no DC flux component in steady state, but slight differences in switch timing and component characteristics can produce a small offset, so the core is sized with margin below its saturation limit.

What should be specified when sourcing an isolated converter transformer?

Topology, switching frequency and duty cycle range, input and output voltage and power, whether a gap is required, peak flux density with margin, core loss at the operating point, isolation and creepage distances, and the expected temperature rise.