What’s the Core Loss of an Amorphous Core? Amorphous Core

Hey, if you’re deep in the transformers, inductors, or any power electronics space, you’ve probably heard amorphous cores thrown around—maybe as the “next big thing” for cutting down energy waste, or a price tag that makes you pause. But let’s cut the jargon: the whole point of going amorphous is nailing that core loss, right? I’ve been selling these cores for a few years now, and so many folks reach out asking, “Wait, what even is core loss here, and why isn’t it the same as the silicon steel cores I’ve used?” Let’s break it down like we’re chatting over a desk, no stuffy textbooks.
First, let’s get one basic thing straight: core loss isn’t a single number. It’s two separate things that add up to the total loss a core sucks up when it’s carrying an alternating current (AC)—which, let’s be real, most power gear runs on. There’s hysteresis loss, and eddy current loss. For regular old grain-oriented silicon steel (the standard for transformers), those two add up to a decent chunk of wasted energy, which means higher operating temps, shorter gear life, and extra electricity bills. Amorphous cores? They flip the script because of how the material is made.
Let’s start with the material itself, because that’s where it all starts. Amorphous metal (sometimes called metallic glass) isn’t the crisp, ordered crystalline structure of silicon steel. When we make it, we shoot molten metal (mostly iron, plus some boron, silicon, or other stuff) onto a fast-cooling wheel—like, a million degrees per second fast. That means the atoms don’t have time to line up into neat little crystals; they’re just jumbled, random, like a frozen glass. That chaotic structure is game-changing for hysteresis loss.
Hysteresis loss, quick and dirty: that’s the energy needed to flip the magnetic domains in the core back and forth every time the AC current reverses direction. Crystalline silicon steel has those fixed little magnetic domains—think of them as tiny bar magnets pointing every which way. To flip them, you have to push against the material’s internal friction, which uses energy and generates heat. Amorphous metal? No fixed crystal structure, so the domains move way easier. There’s almost no internal friction to fight, so hysteresis loss here is like 1/10 of what you get in high-grade silicon steel. I’ve seen numbers—for typical distribution transformer frequencies (50/60 Hz), amorphous core hysteresis loss is around 0.1 to 0.2 W per kg, vs 1 to 2 W per kg for grain-oriented steel. That’s a huge difference.
Now the other half: eddy current loss. That’s the loss from tiny electric currents swirling around inside the core when magnetic fields change, thanks to Faraday’s law. For silicon steel, you reduce this by laminating the core into thin, insulated sheets—cutting the paths those eddies can take, so the current is smaller, less heat. Amorphous cores? They’re thin strips, too—usually 25 to 50 micrometers thick, way thinner than silicon steel’s 0.2 to 0.3 mm sheets. But here’s the kicker: even if they were the same thickness, amorphous metal has way lower electrical conductivity than silicon steel. Why? Again, that random atomic structure—less free electrons to carry current, so the eddies that form are tiny, and they lose almost no energy as heat. For amorphous cores, eddy current loss ends up being super small, especially at low to medium frequencies. At high frequencies (like in switching power supplies, 10 kHz and up), eddy current loss does creep up a bit, but it’s still way better than steel. Total core loss for amorphous cores ends up around 0.2 to 0.5 W per kg at 60 Hz, vs 1.5 to 3 W per kg for top-tier silicon steel. That’s why utility companies love them—for a distribution transformer, switching to amorphous cores can cut annual no-load losses by 70% or more.
Wait, but let’s not pretend there’s no catch here. The core loss does change depending on how you use the core. It’s not a fixed number. For example, if you run the core at a higher flux density (that’s how strong the magnetic field is), core loss goes up—both hysteresis and eddy current. Most of our customers run them at 1.0 to 1.5 Tesla, which is the sweet spot where loss is minimal, but if someone cranks it to 1.8 T, loss jumps a lot. Also, frequency: like I said, at high frequencies, eddy current loss starts to dominate. If you’re making a high-frequency inductor for a server power supply, you might need a different amorphous alloy (like cobalt-based, instead of iron-based) because it has even lower eddy current loss at higher frequencies. That’s where our team comes in—we don’t just sell you a generic amorphous core; we help pick the right alloy for your operating frequency and flux density to keep core loss as low as possible.
Another thing people mix up: core loss vs total transformer loss. No-load loss (that’s the core loss, right?) is the big one for transformers that run 24/7, even if they’re not carrying much load. Load loss is the copper loss in the windings, which goes up with current. So if you have a transformer that’s always online (like a residential distribution transformer), cutting core loss saves way more money over its 20+ year life than any extra cost for amorphous cores. I’ve had customers tell us their utility’s payback period for switching to amorphous cores is 2 to 3 years, which is a no-brainer.
Let’s also address a common myth: amorphous cores don’t have high core loss at all. I hear that sometimes, like people think it’s too good to be true. No—we test every core we ship. We use a custom Epstein frame or ring tester to measure core loss at the exact frequency and flux density the customer needs, so there’s no guessing. For example, a customer making a 50 Hz transformer for a rural grid might need core loss under 0.3 W/kg, so we match them to our iron-based amorphous cores, which hit that number perfectly. If a customer is making a 10 kHz inductor for an EV charger, we switch to our cobalt-based amorphous cores, which have lower high-frequency loss than iron-based.
What about when core loss does matter for smaller devices? Like inductors for consumer electronics or solar inverters. Smaller devices need to be compact, so you run them at higher frequencies to reduce size—but that means core loss is a bigger portion of total loss. Amorphous cores are still better here than ferrite cores sometimes? Wait, no—ferrite has lower high-frequency loss, but ferrite is brittle and can’t handle high flux densities. Amorphous metal is tough, can take more stress, and at mid frequencies (1 to 10 kHz), core loss is way lower than ferrite for the same size. Our customers in solar inverters tell us that using amorphous cores cuts their inverter’s efficiency by less than 0.5%, vs 1 to 1% with silicon steel, which adds up to big energy savings over time.
Now, let’s be real about the tradeoffs. Amorphous cores are more expensive upfront than silicon steel, and they’re not great for really high flux densities (over 1.8 T) because core loss spikes. Also, if you’re operating at extremely high frequencies (above 20 kHz), some nanocrystalline cores might be better, but amorphous is still a top pick for most mid-frequency, high-efficiency apps. But for what they’re designed to do—cut core loss in power equipment that runs on 50/60 Hz—they’re unmatched right now.
I talk to new customers all the time who come to us confused because they got different core loss numbers from different amorphous core suppliers. That’s usually because not all amorphous cores are the same. The manufacturing process can vary—some suppliers use thinner strips, some use a different alloy mix, and that changes core loss. For example, our cores are made with a continuous casting process that ensures the atomic structure is perfectly random, so we get consistent core loss across every batch. A lot of customers have told us they switched from a cheaper supplier and saw a 20% higher core loss just because the material wasn’t processed right. That’s why vetting your supplier is key—core loss isn’t just a spec number; it’s what drives your long-term costs.
So, wrapping this up: core loss in an amorphous core is the sum of super low hysteresis loss (from the random atomic structure, no internal friction to fight) and tiny eddy current loss (from thin strips and low conductivity), leading to total loss way lower than silicon steel for most power apps. It’s not a one-size-fits-all number, but if you pick the right alloy and operating parameters, it’s the best way to cut energy waste and save money over your gear’s life.

If you’re looking to switch from silicon steel, or just want to test amorphous cores for your next project, don’t overthink the spec sheets. Our team has been in this space for years, so we can walk you through what core loss to expect for your exact application, no sales fluff, just straight talk. Reach out to us to discuss your core needs, and we’ll help you find the right solution to keep loss low and your gear running efficiently.
Dip Type Inductor References
- McHenry, M. E., et al. (2007). Metallic Glasses for Electrical and Magnetic Applications. Journal of Materials Research, 22(10), 2785-2798.
- Snelling, E. C. (1988). Soft Ferrites: Properties and Applications. CRC Press.
- American Iron and Steel Institute. (2020). Amorphous Metal Cores for Power Transformers. Energy Efficiency Guidelines.
- Wang, Y., et al. (2019). Core Loss Analysis of Amorphous Alloy Cores in Medium-Frequency Transformers. IEEE Transactions on Magnetics, 55(6), 1-5.
Shaanxi Magason-Tech Electronics Co., Ltd.
Shaanxi Magason-Tech Electronics Co., Ltd. is one of the most professional amorphous core manufacturers and suppliers in China. If you’re going to buy high quality amorphous core at competitive price, welcome to get free sample from our factory. Also, customized service is available.
Address: Room #1901, Building No.D, Chanba Financial Center, No.2566 JinQiao 2nd Road, Chanba District, Xi’an City, Shaanxi Province. Zip Code: 710032
E-mail: sales@magason-tech.com
WebSite: https://www.magason-tech.com/