In my last blog, “Power Electronics. Up, Up & Away!”, I spoke about how resonant topologies and fast GaN power ICs enable high-frequency, high efficiency and high density power supply designs. However, there’s one critical component that can make all the difference … magnetics.
The only one way that high frequency can translate in to higher density is if the magnetic components can operate at higher frequencies while shrinking their size & cost and improving their efficiencies at the same time !! Is this possible? Let’s explore that critical question…
Magnetics Densities & Efficiencies at High Frequencies
Magnetics are complex. Even the most experienced power supply designer can struggle to deeply and thoroughly understand all the performance, density, thermal, cost and EMI implications for magnetic design, especially as we push into higher frequencies. Traditional thinking suggests that as you increase frequencies for a given power supply design, efficiencies will decline, creating the same or greater heat dissipation and, as a result, physical size reduction of the converter will be limited. Two key developments challenge that traditional reality:
- GaN power ICs and soft-switching topologies eliminate virtually all frequency-dependent power losses, allowing frequencies to be increased dramatically without degrading efficiencies, and
- Advances in materials and design are enabling higher frequency magnetics to become smaller, lower cost and more efficient.
Magnetics Performance Factors
Quantifying such power loss reductions at higher frequencies is not for the faint of heart. Multiple loss elements such as core loss, winding loss and eddy current losses must be considered – each with their own subtleties. In addition, without consideration, EMI noise might increase at higher switching speed due to the advance of GaN switches. Thermal and cost implications add to the complexity.
MIT and Dartmouth College have completed excellent work in characterizing high-frequency magnetics which has led to the creation of a modified performance metric which accounts for all magnetic-related power loss elements as frequencies increase. These works have been published in Transactions on Power Electronics(1) and demonstrate that new magnetics materials have actually enabled a 100x increase in switching frequency with higher performance over the last twenty years. By transitioning to new core materials, such as the Hitachi ML91S and the Fair-rite 67, the size, cost and power loss can all be reduced dramatically as the industry moves from 100 kHz operation to 1-10 MHz operation in high-voltage power supplies. Such magnetic components are commercially available today and are being incorporated in to many power supply designs targeted for high-volume consumer and mobile markets such as fast mobile chargers, thin LED TVs and other density-driven applications.
Planar Magnetics – Pushing Density, Squeezing Costs and Simplifying Manufacturing
Traditional, handmade and wire-wound bobbin style magnetics represent a big challenge to power supply manufacturers in manufacturability, labor cost, size & weight. When power supply frequencies can be pushed to 500 kHz or higher, it is possible to transition to planar or PCB-based magnetics in which the windings are designed directly in to the PCB. Furthermore, the size and cost of the magnetic core goes down as the frequency increases. All of these cost, efficiency and density benefits have been demonstrated by Navitas and many of their partners in the frequency range of 500 kHz to 1.5 MHz.
Putting it all together
As described in this blog on magnetics, and prior blogs on GaN power ICs and resonant topologies, we have described a new future where high frequencies and high efficiencies can combine to create higher densities and lower costs. In my next blog, we will talk about the market implications of this high-speed revolution in power electronics. Stay tuned…
Note (1): A. J. Hanson, J. A. Belk, S. Lim, C.R. Sullivan, and D. J. Perreault, “Measurements and Performance Factor Comparisons of Magnetic Materials at High Frequency,” IEEE Trans. On Power Electron. Vol. 31, No. 11, Nov. 2016, pp. 7909-7925. (Readers will pay a fee to access the content)