Customized Reactor Lamination for Solar Power Inverter Cores
Oct 21, 2025
Leave a message
Reactor laminations for solar power inverters are the stacked silicon steel cores inside the DC and AC filter reactors of a photovoltaic inverter. They limit current fluctuation, suppress harmonic interference and cut the energy dissipated in the reactor, which is what allows the inverter to convert the direct current from the array into alternating current at grid quality without overheating.
Because an inverter reactor works in a switching field rather than at mains frequency alone, the lamination grade, sheet thickness and cut profile are chosen against a loss budget that extends far above 50 Hz.
Why Solar Inverter Reactors Need Lamination Cores
Reactors in solar inverters, typically DC reactors on the array side and AC filter reactors on the grid side, face two hard conditions at once: the available power moves with irradiance, and the switching devices chop the current at high frequency. A laminated core handles both. The stack provides a defined inductance that stores and releases energy every switching cycle, while the insulated sheet structure keeps eddy loss low enough for the reactor to stay inside its temperature class.
Three Jobs the Lamination Core Performs
Suppressing current ripple - the magnetic circuit smooths the ripple generated during power conversion, especially during pulse width modulation, so the output current waveform stays close to a sine wave and meets the grid harmonic limits.
Reducing high-frequency loss - inverter switching normally runs between 10 kHz and 50 kHz, so the lamination must keep iron loss low in a high-frequency field and avoid the heat that hysteresis and eddy currents would otherwise produce.
Enhancing magnetic saturation resistance - irradiance changes can produce short current surges, and a high saturation flux density keeps the reactor in its linear region so it continues to limit current and protects the switching devices.
How Lamination Breaks Down Eddy Current
Splitting the core into many thin, insulated sheets raises electrical resistance and shrinks the area available for circulating current. Four effects follow:
Higher resistance - the varnish or oxide coating between sheets acts as a barrier to current flow.
Smaller circulation area - each sheet forms a much smaller loop than a solid core, and induced voltage is proportional to loop area, so the driving voltage falls.
Higher total path resistance - many thin parallel paths resist the eddy current far more than one solid block of steel.
Lower power loss - eddy loss falls with the square of the current, so a modest reduction in circulating current removes a large share of the heat.
Silicon Steel Grades for Inverter Reactors
Commercial grades, high induction grades and domain refined grades are all used in inverter reactors, chosen against the loss budget at the design flux density and switching frequency. Core loss is quoted as specific total loss at 1.7 T and 50 Hz, and induction is quoted at a magnetizing field of 800 A/m.
| Type | Grade | Max core loss at 1.7 T / 50 Hz | Lamination factor | Min induction at 800 A/m |
|---|---|---|---|---|
| Common | 23Q110 | 1.10 W/kg | 97.2 % | 1.82 T |
| Common | 23Q120 | 1.20 W/kg | 97.2 % | 1.82 T |
| Common | 27Q120 | 1.20 W/kg | 97.5 % | 1.82 T |
| Common | 30G120 | 1.20 W/kg | 98 % | 1.82 T |
| High induction | B20P085 | 0.85 W/kg | 97 % | 1.86 T |
| High induction | B23P090 | 0.90 W/kg | 97.2 % | 1.88 T |
| High induction | B23P100 | 1.00 W/kg | 97.2 % | 1.88 T |
| High induction | B27P095 | 0.95 W/kg | 97.5 % | 1.89 T |
| High induction | B30P120 | 1.20 W/kg | 98 % | 1.89 T |
| Domain refined | B20R070 | 0.70 W/kg | 97 % | 1.86 T |
| Domain refined | B23R085 | 0.85 W/kg | 97.2 % | 1.88 T |
| Domain refined | B27RK085 | 0.85 W/kg | 97.5 % | 1.89 T |
Strip width for these grades runs from 35 mm to 1200 mm, so a lamination can be produced for a small board-level reactor or for a cabinet-sized filter choke from the same grade family.
Customization for PV Inverter Builds
| Item | Capability |
|---|---|
| Sheet form | slit strip, cut sheet, punched lamination |
| Nominal thickness | 0.23 mm, 0.27 mm, 0.30 mm |
| Cut profile | straight cut, mitred cut, step-lap joint |
| Joint gap in the assembled core | less than 1.0 mm |
| Window height and width precision | plus or minus 1 mm |
| Testing | core loss, exciting current and dimensional checks |
Windings and clamping are matched to the inverter frame, and laminations are deburred so that stacking pressure does not damage the interlayer coating and short the sheets together.
Frequently Asked Questions
Q: Which lamination grade suits an inverter reactor best?
That depends on the loss budget. Conventional 23Q110 and 27Q120 grades cover most reactors, while domain refined grades such as B20R070 are chosen where high-frequency loss and temperature rise are the limiting factors.
Q: Why does switching frequency matter so much?
Eddy loss rises steeply with frequency, and inverters switch between roughly 10 kHz and 50 kHz. A thinner sheet shortens the eddy current path and keeps that loss within the thermal limit of the reactor.
Q: Can the lamination be supplied as a fully assembled core?
Yes. Cut sheet, punched laminations and assembled, clamped cores are all produced, with joint gap held below 1.0 mm and window dimensions to plus or minus 1 mm.
Q: How is the reactor protected against current surges?
By choosing a grade with high saturation flux density, so the core stays linear during the short surges that follow a rapid change in irradiance and does not lose its current limiting function.
Q: Does an air gap help in an inverter reactor?
Where the reactor must store energy and carry a DC component, a controlled gap keeps inductance stable under bias. The gap length is set mechanically during assembly rather than left to the cut finish.
Q: What information is needed to specify reactor laminations?
Grade, thickness and strip width, cut profile, stack height, window and limb dimensions, target inductance, working current with any DC bias, switching frequency and the required core loss limit.