High Permeability Center-Cut Amorphous C-Shape Cores for 8-100 kHz Inductors
Oct 11, 2025
Leave a message
Why Center-Cut Construction Is Used
A center-cut amorphous C-shape core is produced by winding Fe-based amorphous ribbon into a rectangular block and then cutting the block along its centre line, so that the two halves are mirror images with matching cut faces. Compared with a gap created by stacking separate blocks, a centre cut gives a flat, parallel and repeatable gap across the whole flux path, which is the property that makes the permeability of the finished core predictable in volume production.
The reason permeability matters so much is the frequency band. Amorphous ribbon combines a disordered atomic structure with a thickness of only about 0.020-0.030 mm, so both the hysteresis and the eddy-current components of loss stay low. That is what allows a core of this type to replace laminated silicon steel, ferrite or powdered iron in power inductors operating between about 8 kHz and 100 kHz, where the loss per unit volume decides whether the converter can be cooled passively.
Magnetic Behaviour Under Bias
An amorphous core has a rectangular hysteresis loop in the as-cast condition, which is excellent for a transformer but unsuitable for an inductor carrying DC. Introducing a distributed gap shears the loop, dropping the effective permeability and raising the flux at which saturation begins. Three consequences follow:
Inductance becomes stable with current. A sheared loop means the inductance holds a large fraction of its small-signal value up to the working bias current.
Stored energy rises sharply. For a given core volume, the energy that can be stored before saturation is roughly proportional to the gap length, so a strongly gapped core stores far more energy than a closed one.
Loss changes. Gapping usually reduces the hysteresis contribution at the working flux swing, while the fringing flux around the gap can add local copper loss in the winding; a distributed gap keeps fringing to a minimum.
Saturation induction stays near 1.56 T regardless of gapping, and saturation induction together with permeability varies little with temperature over the normal operating band, which simplifies the thermal design of the choke.
Dimension Series and Window Sizes
The builds below are quoted in millimetres and cover the window widths used for filter and PFC chokes in the 1-10 kW class. Dimensions a and d control the cut faces and are held to the tighter band.
| Build a (mm) | b (mm) | c (mm) | d (mm) | e (mm) | f (mm) |
|---|---|---|---|---|---|
| 9 ± 0.5 | 10 | 32.8 | 15 ± 0.5 | 28 ± 1 | 50.8 ± 1.25 |
| 10 ± 0.5 | 11 | 33 | 20 ± 0.5 | 31 ± 1 | 53 ± 2 |
| 11 ± 0.5 | 13 | 30 | 20 ± 0.5 | 35 ± 1 | 52 ± 2 |
| 11 ± 0.5 | 13 | 40 | 20 ± 0.5 | 35 ± 1 | 62 ± 2 |
| 11 ± 0.5 | 13 | 40 | 25 ± 0.5 | 35 ± 1 | 62 ± 2 |
Application Map from 8 kHz to 100 kHz
Loss density, not permeability alone, decides where each frequency band sits in the design space:
8-20 kHz: AC reactor cores in charging inverters and bidirectional chargers. Ripple is high, the DC component is large, and a moderately gapped core is used so that the inductor survives the worst-case load step.
20-40 kHz: AC and DC filter reactors in high-frequency converters and inverters. The gap is set so that peak flux stays below the knee of the sheared loop at maximum ripple.
40-70 kHz: active front end filter inductors and medium-power switching supplies, where a lighter build with a smaller window gives a better loss-to-volume balance.
70-100 kHz: small storage and differential-mode chokes, where the low loss per unit volume of amorphous ribbon is the main reason for choosing it over powdered iron.
Gap Control in Production
Center-cut cores are usually supplied in matched pairs, with the cut faces impregnated or coated before delivery. Two production details are worth writing into the purchase specification. First, the pairs must be kept matched: mixing halves from different cuts adds an uncontrolled gap and shifts inductance by a wide margin. Second, the gap filler must be specified; a thin epoxy or polyimide shim controls both the gap length and the mechanical bond, and its thickness tolerance becomes part of the inductance tolerance. Inductance should be measured after assembly with the specified clamp force, because the permeance of the magnetic circuit depends on the pressure applied to the mating faces.
Testing and Acceptance
Ribbon is specified against GB/T 19345.1-2017 for Fe-based amorphous soft magnetic alloy strip. Finished cut cores are measured as wound specimens under IEC 60404-6, which covers magnetic property measurement of soft metallic and powder materials from 20 Hz to 100 kHz using ring specimens, or under ASTM A927/A927M, the voltmeter-ammeter-wattmeter method for toroidal cores. Acceptance testing normally reports inductance at the working bias current, core loss at one or two defined flux-density and frequency points, and the DC saturation current. Recording the test point is essential: a loss figure quoted without flux density and frequency is not a specification.
Frequently Asked Questions
Q: What does center cut mean in an amorphous C core?
A: The wound ribbon block is cut once along its centre line so the two halves present flat, parallel and matched faces; the resulting gap is uniform across the flux path and repeatable from unit to unit
Q: What permeability can be expected after gapping?
A: Effective permeability falls as gap length increases, typically from several thousand for a closed cut core to a few tens for a heavily gapped inductor core, and the exact value is measured on the assembled unit
Q: Why are amorphous cores not used at 400 Hz?
A: They can be, but at such low frequencies the loss advantage over laminated silicon steel is smaller and the lower saturation induction means a larger core, so silicon steel is usually the more economical choice
Q: How much DC bias can the core take?
A: Bias capability is set by the gap: a longer gap raises the current at which the core approaches saturation near 1.56 T, and the working point is chosen so that peak flux stays below the knee of the sheared loop
Q: Is the ribbon brittle after winding?
A: The ribbon is ductile as cast but becomes brittle after stress-relief annealing, which is why it is wound into the final toroid or block shape before any heat treatment and handled with edge protection afterwards
Q: Can the two halves be bonded rather than clamped?
A: Yes. Bonding with a controlled-thickness adhesive gives a fixed, non-adjustable inductance and is standard for filter chokes, while clamping is used where inductance must be trimmed after winding

