Flyback transformer leakage inductance stores energy that is not coupled directly to the secondary when the primary switch turns off. That energy produces drain-voltage overshoot, clamp dissipation, ringing, and EMI. The useful engineering target is therefore not simply 鈥渁s low as possible,鈥?but a controlled value compatible with insulation, winding capacitance, voltage stress, efficiency, and production repeatability.
Why leakage energy appears at turn-off
During the primary on-time, current flows through both magnetizing inductance and leakage inductance. Magnetizing energy is transferred through the coupled flux path after switch turn-off. Leakage flux links only part of the winding structure, so its current needs another path. Device capacitance, layout inductance, and the clamp network determine the resulting transient.
Leakage energy increases approximately with leakage inductance and peak current squared. Startup, current limit, low-line full load, and overload may therefore produce more clamp stress than nominal operation.
Measure leakage inductance consistently
A leakage value has little meaning without a method. State which secondary and auxiliary windings are shorted, which winding is measured, test frequency, signal level, fixture compensation, and core assembly condition. Lead and fixture inductance can be significant when the target is small.
Customer and transformer manufacturer should use equivalent connections. A production tester operating at one frequency can be correlated with a development LCR meter, but the relationship should be documented rather than assumed.
Balance leakage against winding capacitance
Interleaving primary and secondary sections generally reduces leakage by improving coupling. It also increases overlapping electric-field area and can raise primary-secondary capacitance. That capacitance conducts common-mode current and may worsen EMI or output-to-earth leakage. Wider margins and sectional bobbins can help insulation or reduce overlap while increasing leakage and conductor length.
The winding arrangement should be selected from measured converter priorities: drain stress, clamp power, common-mode current, dielectric construction, temperature, and available window area.
Choose the clamp from the real waveform
RCD clamps, TVS clamps, active clamps, and regenerative networks handle leakage energy differently. Clamp design should use measured leakage, peak current, switching frequency, input range, device rating, layout parasitics, and the maximum acceptable drain voltage. Component tolerances and temperature alter the final clamp level.
A clamp can limit the main overshoot yet leave high-frequency ringing caused by smaller parasitic inductances and capacitances. Probe technique matters: a long oscilloscope ground lead can exaggerate the observed spike.
Prototype and production checklist
- Leakage inductance with a controlled shorting and fixture method
- Magnetizing inductance, turns ratio, polarity, and winding resistance
- Drain voltage during startup, full load, current limit, and overload
- Clamp voltage, current, power, and component temperature
- Primary-secondary capacitance and common-mode EMI
- Core and winding temperature rise
- Winding placement, margins, insulation, and termination consistency
BaoHui Tech can review high frequency transformer winding options alongside converter waveforms and clamp limits. Sharing the schematic, switching frequency, peak-current limit, insulation requirement, and mechanical envelope allows the transformer manufacturer to propose a measurable leakage range rather than an isolated nominal number.
Frequently asked questions
Can leakage inductance be reduced to zero?
No. Real windings always have some uncoupled flux. The practical goal is a controlled value that supports the complete converter design.
Why does drain ringing remain after adding a clamp?
Device capacitance, PCB layout inductance, leakage distribution, rectifier commutation, and clamp dynamics can create additional resonances.
Should every transformer be tested for leakage?
Often yes when leakage is a critical characteristic and the measurement is practical. The drawing must define the test method and limits.