Engineering evidence, structured

Data that behaves like part of the design.

Reusable device and magnetics data for component selection, loss modelling and AI-assisted engineering workflows — with the conditions and provenance kept attached.

Two connected libraries

From switching event to magnetic component.

The electrical and magnetic records are designed to meet inside the same converter calculation, rather than living as disconnected datasheet summaries.

Why the structure matters

Built for calculation,
not just browsing.

A datasheet is written for a person reading one part at a time. These records are written for a solver comparing thousands of operating points.

01

Curves over headline numbers

Coss(V), Pv(f, B̂, T), µ(Hdc) and Zth(t) preserve the behaviour that decides whether a design closes.

02

Conditions stay attached

Bias, temperature, gate resistance, loop inductance and excitation waveform travel with every value, so comparisons remain meaningful.

03

Sources remain traceable

Digitised and measured records retain their source, revision, fixture and uncertainty for later review.

04

Models close the loop

Switching energy, core loss and thermal networks solve together against an operating or mission profile.

05

Machine-readable by default

Structured curves and surfaces feed scripts, optimisers and design agents without another transcription step.

06

Expandable through measurement

Missing parts can be characterised using the same schema, keeping private and public engineering workflows aligned.

Extend the library

Need a device or core we do not list?

We characterise semiconductor and magnetic components to order, then return the data in the same reusable, traceable format.