Flyback Transformer Air-Gap Design — Chapter 8

Chapter 8 converts the turns and magnetizing-inductance targets from Chapter 7 into the intentional magnetic reluctance required by a practical flyback transformer. The air gap is not a finishing adjustment: it is a primary design variable that controls magnetizing inductance, energy storage, current capability, fringing fields, tolerance, winding loss, and production repeatability.

← Previous Chapter

Chapter 7 — Primary Turns, Secondary Turns, and Flux Density

Next Chapter →

Chapter 9 — Core Geometry and Magnetic Material Selection

In This Chapter — Click to Expand

1. Why a Flyback Transformer Needs an Intentional Air Gap

A flyback transformer is designed to store magnetic energy during the primary switch ON interval. For most ferrite flyback designs, an ungapped core has far too much effective permeability to support the required magnetizing current and energy storage with a practical combination of turns and inductance.

Adding an intentional air gap increases the magnetic circuit reluctance. This reduces effective permeability, lowers the inductance produced by a given number of turns, and allows the transformer to support greater magnetizing ampere-turns before the ferrite reaches an unacceptable flux condition.

The gap does not increase the ferrite material’s intrinsic saturation flux density. Instead, it changes the relationship among turns, current, magnetizing inductance, and magnetic flux so that a useful energy-storage design can be realized.

L_m=\frac{N_P^2}{\mathcal{R}_m}

where L_m is the primary-referred magnetizing inductance, N_P is the primary turns count, and 𝓡_m is the total magnetic reluctance.

SOLIDMAG ENGINEERING INSIGHT

The Gap Is Part of the Transformer’s Electrical Design

The air gap is not simply a mechanical feature added after the transformer calculations are complete. It is one of the elements that determines magnetizing inductance, peak current, stored energy, fringing loss, and production tolerance.

A gap that produces the desired nominal inductance but causes excessive local winding loss, poor tolerance, or manufacturing instability is not a successful design.

2. Magnetic Reluctance, Core Reluctance, and Gap Reluctance

A simplified magnetic circuit can be represented as the sum of ferrite-core reluctance and air-gap reluctance:

\mathcal{R}_m=\mathcal{R}_{CORE}+\mathcal{R}_{GAP}

For a uniform magnetic section, reluctance is approximately:

\mathcal{R}=\frac{l}{\mu A}

For a simple air gap, ignoring fringing:

\mathcal{R}_{GAP}\approx\frac{l_g}{\mu_0A_g}

Variable	Meaning
𝓡_GAP	Air-gap reluctance
l_g	Effective total gap length
μ₀	Permeability of free space
A_g	Effective gap area

Because ferrite permeability is much higher than the permeability of free space, a physically small air gap can dominate the total reluctance of the magnetic circuit. This is why a fraction of a millimeter can change magnetizing inductance substantially.

ENGINEERING CAUTION

Reluctance Approximations Are Geometry Dependent

Real E-cores, ETD cores, PQ cores, distributed gaps, chamfers, center-leg shapes, and multiple physical gaps do not behave as perfectly uniform magnetic circuits. Use manufacturer A_L data, measured inductance, or field analysis when the simple one-dimensional approximation is no longer adequate.

3. Determine the Required Inductance Factor A_L

Core manufacturers commonly characterize a gapped core with its inductance factor, A_L. This relates winding turns to inductance:

L_m=A_LN_P^2

Therefore the target inductance factor is:

A_{L,TARGET}=\frac{L_m}{N_P^2}

The required A_L value provides an efficient bridge between the electrical design and available factory-gapped core parts. If a manufacturer offers a standard gapped core near the required value, that part can provide better production repeatability than manually establishing the gap during assembly.

Manufacturer A_L values may be listed in nH/turn², µH/turn², or other units. Keep units consistent throughout the calculation.

Design Quantity	Why It Matters
Target L_m	Sets the required inductance
Selected N_P	Transforms inductance into the required A_L
Manufacturer A_L tolerance	Sets part of the production inductance tolerance
Gap construction	Influences realized A_L, fringing, and repeatability

4. Estimate the Effective Gap Length

When gap reluctance dominates the magnetic circuit, magnetizing inductance can be approximated by:

L_m\approx\frac{\mu_0N_P^2A_e}{l_g}

Solving for the estimated total effective gap length:

l_{g,EST}\approx\frac{\mu_0N_P^2A_e}{L_m}

This relationship is valuable for first-pass sizing, but it should not be treated as an exact machining dimension. The actual physical gap needed to obtain the target inductance depends on fringing, core reluctance, effective area, gap location, distributed-gap geometry, and manufacturer-specific construction.

DESIGN ASSUMPTION

Gap-Dominated First-Pass Model

The simple gap equation assumes that the intentional gap dominates the total magnetic reluctance and that the effective gap area is approximately equal to the core effective area. It is most useful as an initial estimate before selecting an actual core or factory A_L value.

5. Stored Energy and the Gap Region

The energy stored in the magnetizing inductance is:

E=\frac{1}{2}L_mI^2

For a low-permeability region such as an air gap, the magnetic-field energy density is approximately:

w_g\approx\frac{B_g^2}{2\mu_0}

A first-order estimate of energy associated with a gap volume is therefore:

E_g\approx\frac{B_g^2}{2\mu_0}A_gl_g

These relationships help explain why an intentionally gapped ferrite structure can store substantially more useful energy than the same ferrite core operated without the gap. The ferrite provides the low-reluctance magnetic path, while the intentional gap creates much of the total reluctance and stores much of the incremental magnetic-field energy.

SOLIDMAG ENGINEERING INSIGHT

Energy Storage Is a Core-and-Gap Function

The ferrite core and air gap should not be evaluated as separate components. The core determines magnetic path, area, window, material loss, and thermal geometry; the gap determines much of the reluctance and energy-storage behavior.

The completed magnetic assembly—not the ferrite alone—is the energy-storage component.

Cross-sectional flyback E-core magnetic circuit showing the center-leg winding package, primary and secondary windings, insulation barrier, intentional center-leg air gap, and magnetic flux path.

Figure 8-1. Simplified E-core magnetic circuit showing a center-leg winding package, intentional center-leg gap, magnetic path, and the relationship among turns, gap, flux, and stored energy.

6. Center-Leg Gap, Shim Gap, and Distributed Gap Constructions

The calculated effective gap can be realized in several physical ways. Each construction changes fringing, tolerance, winding exposure, mechanical assembly, and manufacturing repeatability.

Factory-Gapped Center Leg

A core can be supplied with the center leg ground to provide the required magnetic gap while the outer legs mate closely. This localizes much of the intended gap beneath the winding package and can provide a controlled magnetic design.

A concentrated center-leg gap also produces a concentrated fringing field near the winding, so conductor position relative to the gap must be evaluated.

Spacer or Shim Between Core Halves

A nonmagnetic spacer inserted between mating core halves creates physical separation at the core interfaces. This technique is convenient during prototyping, but the resulting physical gaps can occur at the center and outer legs rather than only at the center leg.

The resulting fringing field distribution, core assembly pressure, adhesive thickness, spacer tolerance, and EMI implications should be evaluated rather than assuming the shim behaves exactly like a center-leg-only gap.

Distributed Gap

A distributed-gap construction divides the required magnetic reluctance among multiple smaller gaps. This can reduce the peak fringing-field intensity at any one gap and may improve winding loss and thermal behavior.

Distributed-gap implementation depends strongly on the selected core family, manufacturer, material, cost, and production method.

Gap Construction	Potential Advantages	Important Tradeoffs
Factory center-leg gap	Repeatable A_L; localized reluctance; defined part number	Strong local fringing near winding; availability limits
Shim/spacer	Easy prototyping; adjustable thickness	Gaps can appear at multiple legs; compression and assembly tolerance matter
Distributed gap	Lower local fringing intensity; potentially improved winding loss	Specialized core; availability and cost; geometry-specific design

7. Fringing Fields and Winding Loss Near the Gap

Magnetic flux spreads outward around the edges of a physical gap. This fringing field can intersect nearby conductors and induce additional eddy-current and proximity loss.

The problem becomes more important when:

The gap is relatively large
The winding is close to the gap
Wide foil or thick conductors are used
Switching frequency is high
Several conductive layers occupy the gap region
Peak current and local field strength are high

Possible mitigation methods include:

Increase winding distance from the concentrated gap
Use a bobbin or winding geometry that provides local spacing
Distribute the gap where practical
Avoid placing wide conductive surfaces immediately adjacent to the gap
Rearrange winding layers
Use field or winding-loss analysis when the local geometry is critical

ENGINEERING CAUTION

Gap Fringing Is a Winding-Loss Mechanism

The magnetic field originates at the gap, but the resulting eddy-current loss occurs in nearby conductors. Track gap-fringing loss as part of the winding-loss and thermal design rather than treating it only as a core-loss term.

Comparison of concentrated center-leg, shim/spacer, and distributed-gap flyback transformer constructions showing representative magnetic fringing-field regions.

Figure 8-2. Comparison of concentrated center-leg gap, shim/spacer gap, and distributed-gap constructions with representative fringing-field regions.

8. Gap Tolerance and Magnetizing-Inductance Tolerance

When the air gap dominates the magnetic reluctance, small gap-length changes can produce meaningful changes in magnetizing inductance. In the gap-dominated approximation:

L_m\propto\frac{1}{l_g}

A larger physical gap therefore lowers inductance, while a smaller gap raises it. The converter consequences can be significant because magnetizing inductance controls current ramp:

\Delta I_P=\frac{V_{P,ON}D}{L_mf_s}

Lower-than-nominal inductance can increase peak current and reduce current-limit margin. Higher-than-nominal inductance can change residual current, operating mode, demagnetization behavior, and transient response.

Production gap variation can result from:

Core grinding tolerance
Spacer thickness
Adhesive thickness
Core-face flatness
Assembly pressure
Core misalignment
Clip or clamp force
Temperature
Mechanical stress

SOLIDMAG ENGINEERING INSIGHT

Specify Inductance Tolerance from Converter Limits

Do not choose an air-gap tolerance simply because a grinding process or spacer supplier can hold it. Work backward from the allowable magnetizing-inductance range and the resulting current, operating-mode, flux, and control margins.

The magnetic manufacturing tolerance should be derived from the electrical system requirement.

9. Recalculate Current and Flux After Selecting the Gap

Once a practical gap or A_L value has been selected, recalculate the realized magnetizing inductance:

L_{m,ACT}=A_{L,ACT}N_P^2

Then update the primary current ramp:

\Delta I_P=\frac{V_{P,ON}D}{L_{m,ACT}f_s}

and the corresponding flux excursion:

\Delta B=\frac{L_{m,ACT}\Delta I_P}{N_PA_e}

For DCM or BCM operation beginning near zero magnetizing current:

B_{PK}\approx B_{START}+\frac{L_{m,ACT}I_{P,PK}}{N_PA_e}

For CCM, the nonzero minimum current must also be included when evaluating magnetic bias and peak flux.

Perform this recalculation at minimum, nominal, and maximum expected magnetizing inductance. A nominal gap that produces acceptable results is insufficient if one tolerance extreme violates peak-current, saturation, or control constraints.

10. Measuring the Realized Gap and Inductance

In production, the most meaningful quantity is usually the completed assembly’s magnetizing inductance rather than the physical gap dimension alone.

A first verification can be performed by measuring the primary inductance with the secondary and auxiliary windings open, using a defined test frequency and signal level.

Document:

Test frequency
Test signal amplitude
Core assembly condition
Temperature
Open or shorted state of other windings
Instrument and fixture
Acceptance limits

Small-signal inductance does not verify saturation capability at operating current. Where appropriate, also measure or evaluate:

Inductance under DC bias
Current-ramp behavior in the converter
Saturation onset
Leakage inductance
Core and winding temperature
Production repeatability

The physical gap may still be inspected for process control, but electrical acceptance should be tied to the transformer quantities that matter to converter operation.

11. Worked Example — Air Gap for the Chapter 7 Transformer

DESIGN ASSUMPTION

Educational Continuation of the Chapter 5–7 Design

This example uses the provisional Chapter 7 values to estimate the required A_L and effective gap. It does not select a real production core part or account for exact manufacturer fringing corrections.

Parameter	Value
Primary turns	32
Target magnetizing inductance	147.3 µH
Candidate effective core area	80 mm²
Peak primary current	3.06 A
Flux excursion	0.176 T
Stored energy at peak current	Approximately 690 µJ

Step 1 — Calculate the Target A_L

A_{L,TARGET}=\frac{L_m}{N_P^2}

A_{L,TARGET}=\frac{147.3\ \mu\mathrm{H}}{32^2}

A_{L,TARGET}\approx143.8\ \mathrm{nH/turn^2}

A practical core candidate should therefore provide an A_L value near 144 nH/turn² after gapping.

Step 2 — Estimate the Total Effective Gap

l_{g,EST}\approx\frac{\mu_0N_P^2A_e}{L_m}

l_{g,EST}\approx\frac{(4\pi\times10^{-7})(32^2)(80\times10^{-6})}{147.3\times10^{-6}}

l_{g,EST}\approx0.70\ \mathrm{mm}

This is the estimated total effective gap in the gap-dominated approximation. A manufacturer’s factory-gapped A_L data should take precedence over treating 0.70 mm as an exact machining instruction.

Step 3 — Verify Peak Stored Energy

E_{PK}=\frac{1}{2}L_mI_{P,PK}^2

E_{PK}=\frac{1}{2}(147.3\times10^{-6})(3.06)^2

E_{PK}\approx690\ \mu\mathrm{J}

Step 4 — Verify Flux from Current

\Delta B=\frac{L_mI_{P,PK}}{N_PA_e}

\Delta B\approx0.176\ \mathrm{T}

The gap, inductance, current, turns, and flux values are therefore internally consistent under the stated first-pass assumptions.

Step 5 — Evaluate a ±10% A_L Example

If the realized A_L varies by ±10%, the magnetizing inductance would vary approximately by the same percentage because the turns count is fixed:

L_{m,MIN}=0.90(147.3\ \mu\mathrm{H})\approx132.6\ \mu\mathrm{H}

L_{m,MAX}=1.10(147.3\ \mu\mathrm{H})\approx162.0\ \mu\mathrm{H}

At the same 100 V, 45% duty cycle, and 100 kHz switching frequency, the minimum-inductance current ramp becomes:

I_{P,PK,MINL}\approx\frac{100(0.45)}{(132.6\times10^{-6})(100000)}\approx3.39\ \mathrm{A}

The tolerance therefore increases the nominal 3.06 A current to approximately 3.39 A before controller and frequency tolerances are added. This demonstrates why gap and A_L tolerance must be included in current-limit analysis.

Quantity	First-Pass Result
Target A_L	143.8 nH/turn²
Estimated total effective gap	0.70 mm
Peak stored energy	690 µJ
Nominal flux excursion	0.176 T
−10% inductance example	132.6 µH
+10% inductance example	162.0 µH
Peak current at −10% inductance	3.39 A

SOLIDMAG ENGINEERING INSIGHT

The Gap Calculation Is Only the Beginning of Gap Design

The simple calculation produces a target A_L and an estimated effective gap. The production design still requires an actual core family, factory or machined gap option, tolerance, fringing evaluation, winding placement, thermal analysis, and measurement.

A physically accurate gap specification must come from the selected core geometry and manufacturing process.

Flyback transformer worked gap-design example showing target AL, estimated center-leg air gap, stored magnetic energy, and the effect of magnetizing-inductance tolerance on peak current.

Figure 8-3. Chapter 7 worked design converted into target A_L, estimated effective gap, stored energy, and inductance-tolerance effects on peak current.

12. Air-Gap Design Checklist

Primary turns and target magnetizing inductance are finalized enough for gap iteration.
Target A_L is calculated from L_m/N_P².
A manufacturer gapped-core option has been checked before specifying custom grinding.
Estimated effective gap has been calculated with appropriate assumptions.
Center-leg, shim, or distributed-gap construction has been selected deliberately.
Fringing fields near the winding have been evaluated.
Gap and A_L tolerances are included in minimum/maximum inductance analysis.
Peak current is recalculated at minimum inductance.
Flux density is recalculated after the realized inductance is known.
Core loss and saturation margin remain acceptable.
The completed assembly’s magnetizing inductance will be measured under defined test conditions.
Production acceptance limits are tied to converter requirements.

SOLIDMAG ENGINEERING INSIGHT

Specify What the Converter Needs, Then Control the Manufacturing Process

The most useful production specification is not simply “gap = 0.70 mm.” It is a controlled magnetic assembly that delivers the required magnetizing-inductance range, current capability, flux margin, fringing behavior, and repeatability.

Use physical gap dimensions as manufacturing controls where appropriate, but validate the finished transformer electrically.

13. Design Handoff to Core Geometry and Material Selection

The Chapter 8 calculations establish the required reluctance and first-pass gap behavior, but they do not yet prove that the provisional core geometry and ferrite material are optimal.

Chapter 9 will evaluate:

Core effective area and minimum cross-sectional area
Winding-window area
Area product
Mean turn length
Effective core volume
Core geometry families
Bobbin availability
Power-ferrite material options
Frequency and temperature dependence
Core loss
Saturation margin
Mechanical and manufacturing constraints

The core and material selection may force turns, gap, conductor, and winding calculations to be repeated. That iteration is expected—the magnetic design is converging toward a buildable component.

SOLIDMAG ENGINEERING INSIGHT

The Gap Cannot Be Final Until the Core Is Final

Gap length, A_L, fringing, core area, and stored energy all depend on the actual core geometry. Chapter 8 therefore establishes the magnetic-reluctance requirement; Chapter 9 will determine which physical core and material can satisfy it most effectively.

Technical References for This Chapter

Use the selected core manufacturer’s datasheet and gapped-core data for final A_L, gap, effective area, tolerance, and material information. The equations in this chapter are first-pass magnetic-circuit relationships and should be verified against actual core geometry and measured transformer inductance.

Chapter 7 — Primary Turns, Secondary Turns, and Flux Density

Complete Flyback Transformer Design Guide

Chapter 9 — Core Geometry and Magnetic Material Selection

Flyback Transformer Designer

Ready to Design a Flyback Transformer?

Move from design theory to an automated magnetic design. Enter the converter requirements and evaluate a coordinated transformer candidate, winding geometry, loss estimates, thermal performance, and CAD output.

Start Flyback Transformer Design →

Explore Magnetic Design Automation →

← Previous Chapter

Chapter 7 — Primary Turns, Secondary Turns, and Flux Density

Next Chapter →

Chapter 9 — Core Geometry and Magnetic Material Selection

Flyback Transformer Air-Gap Design — Chapter 8

Table of Contents

1. Why a Flyback Transformer Needs an Intentional Air Gap

The Gap Is Part of the Transformer’s Electrical Design

2. Magnetic Reluctance, Core Reluctance, and Gap Reluctance

Reluctance Approximations Are Geometry Dependent

3. Determine the Required Inductance Factor A_L