Magnetic potential: advanced characterisation for power electronics

One part of CSA Catapult’s characterisation capability is the BH Loop Analyser, which monitors core losses, whilst also facilitating collaborative research.

Dr. Gerardo Calderon-Lopez, principal power electronics engineer at the Catapult explained: “The system allows us to measure the core losses of a wide range of magnetic cores with high accuracy across a range of frequencies and voltage excitations. By measuring the B-H curve (or hysteresis loop) we can characterise and model the that power convertors depend on.”

Power switches based on compound semiconductors enable higher switching frequencies (in comparison with more traditional materials), but this comes with trade-offs; often affecting thermal performance or reliability. The team therefore need to understand how the magnetic devices perform at these higher frequencies.

This is done by measuring the losses produced by the flow of magnetic flux through the soft magnetic cores.

Almost any topology in power electronics uses a magnetic device to perform power conversion or power conditioning. Analysing the BH loop and other characteristics is useful to power electronics designers when they are selecting a magnetic material, the core dimensions and a core shape– enabling them to design optimal inductors or transformers.

The capabilities the Catapult can bring to bear in magnetics also help to de-risk – investing in this type of equipment would be a prohibitive cost for many manufacturers.

The wide range of measurements available is also a unique capability, with the BH Loop Analyser able to go up to 150 V, 6A AC current, 30 A DC current, and up to 10 MHz in frequency with sinusoidal waveforms.

It can also test under square voltage waveforms with variable duty cycle. This feature is important because several power converters rely on this type of excitation.

Gerardo explained that further innovations are in the pipeline at the Innovation Centre: “Next, we will develop the capacity to measure at higher temperatures, when the magnetic parameters change. We’d also like to test at extremely low temperatures.

“We’re working with a researcher from the University of Bristol, Dr Jun Wang, for developing loss models using artificial intelligence techniques. This has the potential to cut product development time even further, allowing partners to select models quickly for their applications.”