The Medium-Frequency Transformer
- MFT operating frequency 5-100 kHz, commercial sweet spot 10-30 kHz
- Conventional silicon-steel cores fail above ~1 kHz due to eddy-current and hysteresis losses
- Nanocrystalline cores (VITROPERM, Finemet) preferred: 1.2 T saturation, low losses at MFT range
- Demonstrated lab efficiency 99.41% on 1 kVA / 5 kHz nanocrystalline MFT-DAB at 0.9 T
- Thermal management: forced-air, deionized-water, or dielectric-fluid immersion — largest subsystem on MV-SST products
The medium-frequency transformer (MFT) is the magnetic component at the heart of every solid-state transformer (SST). The MFT provides galvanic isolation between the input and output sides of the SST and provides the voltage step-down ratio that the silicon-carbide power devices alone cannot deliver economically. Although the SST is named for its semiconductor switches, the actual engineering challenge of building a commercial SST is overwhelmingly concentrated in the MFT design.
The operating frequency range for commercial SST MFTs sits between 5 kilohertz and 100 kilohertz, depending on power rating and topology choice. Higher frequency means smaller magnetic volume and lower copper conductor mass; the trade-off is higher switching losses in the power semiconductors and higher dielectric losses in the magnetic core material. The sweet spot for most medium-voltage SSTs lands between 10 and 30 kilohertz.
The magnetic core material is the second design choice that determines MFT performance. Conventional silicon-steel cores, which dominate the 50 and 60 hertz iron-core transformer industry, do not work effectively above approximately 1 kilohertz because of eddy-current and hysteresis losses. Three alternative core materials are used for MFTs: ferrite, amorphous metal alloys, and nanocrystalline alloys. Nanocrystalline cores have emerged as the preferred material for high-power MFTs because they combine high saturation flux density (approximately 1.2 tesla) with low core loss at the 5-100 kilohertz operating range.
The dominant nanocrystalline product family in commercial SST applications is VITROPERM, sold by Vacuumschmelze, although several other suppliers including Hitachi Metals (Finemet) and various Chinese suppliers offer competing nanocrystalline cores. An MFT-DAB (dual-active-bridge) laboratory prototype built at 1 kilovolt-ampere and 5 kilohertz with a flux density of 0.9 tesla has demonstrated efficiency of 99.41 percent. Scaling that efficiency to multi-megawatt commercial SSTs is the active engineering work.
The thermal-management challenge in commercial-scale MFTs is also non-trivial. The compact form factor that the MFT enables for the SST also concentrates the residual losses in a small volume. High-power MFTs use forced-air cooling, deionized-water cooling, or in some cases dielectric-fluid immersion. The cooling architecture is the largest physical and capital subsystem on most commercial MV-SST products and is the segment where ABB, Hitachi Energy, Eaton, and the startups differentiate most visibly.
Why It Matters
The nanocrystalline core is not interchangeable with the ferrite and amorphous materials it competes against — the 1.2 tesla saturation flux density and low loss at 5-100 kilohertz are what make 10-30 kilohertz operation viable for medium-voltage SST designs. Without a reliable supply of VITROPERM-class nanocrystalline material from Vacuumschmelze, Hitachi Metals (Finemet), or comparable Chinese sources, the SST roadmaps of ABB, Hitachi Energy, Eaton, and the startup wave cannot be commercialized at scale. The 99.41 percent efficiency demonstrated in laboratory DAB prototypes at 1 kilovolt-ampere depends directly on achieving low core loss in this frequency range, and scaling that result to the multi-megawatt class is the central unresolved engineering challenge in the SST supply chain.
Critical Perspective
Editorial correction: This post is part of MGRID’s Solid-State Transformer industry coverage. As of May 2026, that body of work systematically framed manufacturer announcements, funding rounds, and laboratory demonstrations as commercial deployments. The reality is that field-deployed commercial-class SST in revenue service globally is measured in single digits, and almost every product cited in this series is at “announced” or “funded” stage, not “operational.” Readers should treat the specific claims in this post against the standards documented in our [SST Industry Reality Check](/?p=9044) (the per-claim audit table maps marketing language to verifiable deployment status). The corrective article is the canonical reference for SST industry reality; this post remains published with its original framing so the editorial drift is traceable.
Related Coverage
Thermal management: forced-air, deionized-water, or dielectric-fluid immersion — largest subsystem on MV-SST products
