Flyback Transformer Design: Tips for High-Efficiency Isolated Power Supplies

Flyback Transformer Design: Tips for High-Efficiency Isolated Power Supplies

📅 Updated: April 2026 | ⏱ 12 min read | 🔧 Magnetic Design

The flyback converter is one of the most popular topologies for isolated power supplies in the 1W–150W range. Its simplicity, low component count, and ability to generate multiple outputs make it a go‑to choice for auxiliary power supplies, battery chargers, LED drivers, and industrial controls. However, the heart of any flyback converter—the flyback transformer—is also the most critical and often misunderstood component. Poor transformer design leads to excessive leakage inductance, voltage spikes, low efficiency, and even catastrophic failure. In this guide, we’ll share practical tips for designing high‑efficiency flyback transformers for isolated power supplies, covering core selection, winding techniques, leakage inductance control, snubber optimisation, and verification methods.

Why the Flyback Transformer Is Different

Unlike a conventional transformer that transfers energy simultaneously from primary to secondary, a flyback transformer operates as a coupled inductor. It stores energy in its core during the primary switch ON time and releases it to the secondary during the OFF time. This means the core must handle a significant DC bias without saturating, and the winding arrangement must minimise leakage inductance (which causes voltage spikes on the primary MOSFET). Key differences from a standard transformer:

• The core has a gapped structure to store energy (a non‑gapped ferrite core would saturate almost instantly).
• Primary inductance (L_P) is a key design parameter, not just turns ratio.
• Leakage inductance must be tightly controlled to reduce snubber losses.
• Interleaving windings is often used to improve coupling.

💡 Key Insight: A well‑designed flyback transformer can achieve leakage inductance of 1–2% of primary inductance. Poor design can result in 10–20% leakage, drastically reducing efficiency and stressing the MOSFET.

Step 1: Define Electrical Specifications

Before selecting a core, you must define the converter’s operating conditions:

• Input voltage range (V_IN,min, V_IN,nom, V_IN,max)
• Output voltage(s) and current(s)
• Switching frequency (f_SW) – typically 50–250 kHz
• Maximum duty cycle (D_max) – often 0.45–0.5 for CCM, up to 0.7 for DCM
• Target efficiency (η) – 80–90% for initial calculations
• Operating mode: Continuous Conduction Mode (CCM) or Discontinuous Conduction Mode (DCM)

From these, calculate the required turns ratio (n = N_P/N_S), primary inductance (L_P), and peak primary current (I_PK). For a DCM flyback:

L_P = (V_IN,min² × D_max² × η) / (2 × P_out × f_SW)

I_PK = (2 × P_out) / (V_IN,min × D_max × η)

For CCM, the equations are more complex and require setting a ripple current factor (typically 30–50% of I_PK).

Step 2: Choose the Right Core Material and Shape

Ferrite is the universal choice for flyback transformers due to its low cost, high resistivity (low eddy losses), and good high‑frequency characteristics. Common grades:

• PC40 / N87 / 3C90: General purpose, good for 50–200 kHz.
• PC95 / N97: Lower losses, higher frequency (up to 500 kHz).
• Powdered iron: Rarely used in flybacks because of higher core loss.

Core shape influences leakage and winding ease:

• E cores (EE, EER, ETD): Good balance of cost, availability, and window area. Popular for low‑ to medium‑power flybacks.
• PQ cores: Optimised for surface mount, good shielding, low EMI.
• RM cores: High shielding, used in telecom and medical.
• Toroids: Very low leakage, but difficult to gap and wind; not recommended for production flybacks.

Select a core size based on the area product (A_P = A_e × A_w) or using manufacturer’s power handling charts. For example, a 30W flyback might use an EE25, while 100W requires ETD39 or similar.

Step 3: Calculate Primary Turns and Air Gap

To prevent saturation, the core must have an air gap. The required number of primary turns (N_P) is given by:

N_P = (L_P × I_PK) / (B_peak × A_e)

Where B_peak is the peak flux density (typically 0.2–0.3 T for ferrite at 100 kHz). For CCM, use the peak-to-peak flux swing. The air gap length (l_g) can be approximated:

l_g = (μ₀ × N_P² × A_e) / L_P – (l_e / μ_r)

In practice, manufacturers specify the AL value (inductance per turn squared) for gapped cores. Choose a gap that gives the desired L_P with the calculated N_P.

⚠️ Critical: Never use an ungapped ferrite core in a flyback converter. The energy stored in the gap is essential; an ungapped core will saturate even at low power, causing excessive current and likely destroying the MOSFET.

Step 4: Secondary Turns and Multiple Outputs

The secondary turns are derived from the turns ratio (n = N_P/N_S) calculated earlier. For multiple outputs, each secondary winding’s turns are proportional to its desired voltage, accounting for diode drops:

V_OUT,k = (N_S,k / N_P) × V_IN × (D/(1-D)) – V_F,k

In practice, start with the main regulated output (usually the highest power) and derive other windings by scaling. Tight coupling between primary and secondary windings is essential for good cross‑regulation.

Step 5: Minimise Leakage Inductance – Winding Techniques

Leakage inductance is the energy stored in the magnetic field that does not link the secondary. It causes a voltage spike when the primary MOSFET turns off, requiring a snubber and reducing efficiency. To minimise leakage:

• Use interleaving: Sandwich the secondary winding between two halves of the primary winding (e.g., primary – secondary – primary). This dramatically reduces leakage.
• Keep windings tight: Use full bobbin width; avoid random winding.
• Minimise insulation thickness: Use triple‑insulated wire for secondary to reduce creepage without adding thick tape.
• Reduce number of winding layers: More layers increase leakage. Use parallel strands (Litz wire) for high‑frequency currents.
• For multiple outputs, wind all secondaries simultaneously (bifilar) to improve coupling.

A well‑designed flyback transformer should have leakage inductance below 1–2% of L_P. Measure it by shorting all secondary windings and measuring primary inductance.

🔧 Pro Tip: If you cannot achieve low leakage with standard construction, consider a resonant flyback (quasi‑resonant) or active clamp topology, which recycles leakage energy rather than wasting it.

Step 6: Design the RCD Snubber

Even with good coupling, some leakage remains. An RCD snubber across the primary winding clamps the voltage spike and dissipates the leakage energy. Design the snubber for minimal power loss while protecting the MOSFET. Steps:

1. Choose a clamp voltage V_clamp = 1.2–1.5 × V_reflected (where V_reflected = V_out × N_P/N_S).
2. Select snubber capacitor C_sn to limit voltage ripple (typ. 10–47 nF).
3. Calculate snubber resistance: R_sn = (V_clamp²) / (0.5 × L_leak × I_PK² × f_SW).
4. Power dissipation in R_sn ≈ V_clamp² / R_sn. Choose a resistor rated 2–3× that value.

For high efficiency, consider a lossless snubber (e.g., passive clamp using an extra winding and diode) or an active clamp circuit.

Step 7: Wire Gauge and Insulation

• Calculate RMS currents for primary and secondary windings. For DCM, I_RMS,pri = I_PK × √(D/3).
• Select wire gauge to keep current density ≤ 4–6 A/mm² for natural cooling.
• At high frequencies (>100 kHz), skin and proximity effects matter. Use Litz wire or multiple thinner strands.
• For isolated flybacks, ensure sufficient insulation between primary and secondary (e.g., triple‑insulated wire, reinforced insulation tape). Safety standards (IEC 60950, 60601) dictate creepage and clearance distances.

Step 8: Practical Verification and Iteration

After building a prototype, verify:

• Primary inductance – should match design (±10%).
• Leakage inductance – should be <2% of L_P.
• Core saturation: Monitor primary current waveform; if it rises sharply at the end of the ON time, reduce peak current or increase gap.
• Drain voltage spike: Should be within MOSFET’s V_DSS minus margin (at least 20%).
• Temperature rise: Core and windings should stay below 100°C for reliability.

If efficiency is lower than expected, check snubber losses, core material (use lower‑loss grade), and winding DC resistance. Small iterations in turns ratio or gap length often yield significant improvements.

Common Mistakes and How to Avoid Them

• Too few primary turns → core saturation: Always calculate B_peak and verify with a current probe.
• Excessive leakage due to poor winding layout: Use interleaving; avoid leaving empty space on bobbin.
• Underrated snubber resistor: Use a resistor with sufficient power rating (often 2–5W) and pulse‑withstanding capability.
• Ignoring parasitic capacitance: High inter‑winding capacitance causes common‑mode EMI. Use Faraday shield (copper foil between primary and secondary) if needed.
• Incorrect gap measurement: Grinding the centre leg of an E core is the standard method; do not rely on tape gaps (they are unstable).

Conclusion: Master the Flyback Transformer for Efficient Isolated Power

Designing a high‑efficiency flyback transformer is both an art and a science. By carefully selecting core material and geometry, calculating primary turns and air gap, minimising leakage inductance through interleaved winding, and optimising the snubber circuit, you can achieve 85–92% efficiency in a compact, cost‑effective isolated power supply. Always prototype and measure leakage, saturation, and temperature. With the tips provided in this guide, you will be well equipped to design robust flyback transformers for a wide range of isolated power supply applications—from auxiliary power in EV chargers to industrial control systems and LED drivers. © 2026 Power Electronics Guide – Your resource for flyback transformer design, isolated power supplies, and magnetic component engineering.