{"id":3648,"date":"2026-10-21T11:53:35","date_gmt":"2026-10-21T03:53:35","guid":{"rendered":"http:\/\/www.nattonhatban.com\/blog\/?p=3648"},"modified":"2026-10-21T11:53:35","modified_gmt":"2026-10-21T03:53:35","slug":"how-to-reduce-the-emi-of-an-air-core-inductor-481a-f40929","status":"publish","type":"post","link":"http:\/\/www.nattonhatban.com\/blog\/2026\/10\/21\/how-to-reduce-the-emi-of-an-air-core-inductor-481a-f40929\/","title":{"rendered":"How to reduce the EMI of an air core inductor?"},"content":{"rendered":"

What\u2019s up fellow engineers, hobbyists, and anyone who\u2019s ever stared at a wonky circuit board thinking \u201cwhy is this thing glitching for no reason\u201d? If you\u2019re working with air core inductors, you\u2019ve probably run into EMI issues at some point\u2014right? EMI, or electromagnetic interference, is that annoying side effect of inductors doing their job, where they spill electromagnetic energy all over the place and mess up nearby components, or even get messed up themselves. As a guy who\u2019s been supplying air core inductors for, let\u2019s say, over a decade now, I\u2019ve seen every trick in the book for taming this stuff. Today, I\u2019m breaking down what actually works, not the jargon-heavy junk you\u2019ll find in a textbook that never mentions real-world assembly lines or field tests. Air Core Inductor<\/a><\/p>\n

<\/p>\n

First, let\u2019s get one thing straight: air core inductors are great. No core losses, lightweight, perfect for high-frequency applications (RF, power converters, automotive electronics, you name it). But their big flaw is that they\u2019re just a coil of wire in open space\u2014so they radiate EMI way more than core-based inductors. If you\u2019re using them in a design where EMI compliance matters (think FCC Part 15 for consumer gear, or ISO 11452 for cars), cutting that radiation without killing your inductor\u2019s performance is non-negotiable. Let\u2019s start with the basics that most people skip because they think \u201cair core = simple, so I just wind a coil\u201d. Spoiler: simple doesn\u2019t mean effective for EMI.<\/p>\n

1. Coil Geometry: Small Changes, Big EMI Cuts<\/h3>\n

Let\u2019s talk about how you wind the coil first. The shape and spacing of your turns are way more impactful than most folks realize. I used to get calls from a design engineer who was winding his inductors with 1mm spacing between turns because he thought \u201cmore space = less capacitance, better high-frequency performance\u201d. Sure, parasitic capacitance goes down, but that also means each turn acts like a little antenna radiating on its own. Tighten that spacing a bit\u2014wait, not too tight, otherwise you get inter-turn capacitance that causes self-resonance issues. We\u2019ve found that for most standard inductors, a turn spacing of 0.2\u20130.5x the wire diameter hits the sweet spot. That reduces the individual turn\u2019s radiation because the magnetic fields of adjacent turns cancel each other out a little. It\u2019s not magic, it\u2019s just physics: magnetic fields from opposite sides of a coil turn point in opposite directions, so close spacing makes those fields overlap and cancel instead of escaping.<\/p>\n

Then there\u2019s the coil shape. Round coils are common, but octagonal or even rectangular? Wait, no\u2014wait, actually, a toroidal air core? No, wait, air core toroids don\u2019t have a closed magnetic path, so they\u2019re worse for radiation than core-based ones. Wait, no, I\u2019m mixing things up. Let\u2019s correct that: a solenoid coil\u2019s end turns are a huge source of EMI. The magnetic field at the ends of a solenoid leaks out straight instead of looping around inside. So if you make the coil longer and skinnier, or shorter and fatter? Studies (and our in-house testing) show that a shorter, wider coil has less end radiation than a long, skinny one. Wait, let\u2019s test this with a real example: a 10\u00b5H air core inductor for a Bluetooth circuit. We had two prototypes: one 50mm long, 10mm diameter (long\/skinny) and one 20mm long, 25mm diameter (short\/fat). The long\/skinny one had an EMI emission 12dB higher at 2.4GHz, right in the middle of Bluetooth\u2019s band. That\u2019s a massive difference just from changing the aspect ratio. Most people default to aspect ratios of 5:1 or higher, but dropping that to 2:1 or 1.5:1 cuts end radiation so much.<\/p>\n

Also, winding direction: if you\u2019re winding a multi-layer coil, don\u2019t just stack turns straight on top of each other. That\u2019s called \u201cunidirectional winding\u201d, and it makes the magnetic fields from each layer add up instead of canceling. Instead, use \u201cbifilar winding\u201d if you have enough wire, or if you\u2019re doing multi-layer, alternate the winding direction every layer. Wait, bifilar winding is when you wind two wires side by side, right? For inductors, bifilar can reduce EMI because the current in each wire flows opposite directions, so their magnetic fields cancel. But only use that for low-current applications\u2014if you\u2019re running 10A through a bifilar coil, you\u2019ll get too much heat from the opposing currents. We use bifilar exclusively for RF inductors, where current is low and EMI is critical.<\/p>\n

2. Shielding: No More Myth About \u201cYou Can\u2019t Shield Air Core Inductors\u201d<\/h3>\n

A lot of old-timers will tell you \u201cair core inductors can\u2019t be shielded because the shield will add a core, which messes up their high-frequency performance\u201d. That\u2019s half true. You don\u2019t want a solid ferromagnetic shield, that will saturate at high frequencies and add eddy current losses. But a non-magnetic conductive shield? Yes, that works, and we\u2019ve been using it for decades for our high-EMI applications.<\/p>\n

The key here is that the shield has to be a Faraday cage for the magnetic field. Wait, magnetic fields\u2014so a Faraday cage works for electric fields, but for magnetic, you need a conductive shield that\u2019s thick enough to provide skin effect. At high frequencies (above 1MHz, which is where most EMI issues pop up), skin depth is tiny\u2014for copper, it\u2019s about 20\u00b5m at 10MHz, so a 100\u00b5m thick copper foil shield is enough. The shield needs to cover the entire inductor, and it has to have no gaps\u2014if there\u2019s a gap, the magnetic field leaks through exactly like RF through a hole in a metal can. We usually make our shields as a thin copper wrap with a solder seam, and we connect one end to the ground plane on the PCB. Wait, only connect one end\u2014if you connect both ends to ground, you create a shorted turn around the inductor, which will reduce your inductance by 10\u201320%. That\u2019s a mistake a lot of people make, and it kills their inductor\u2019s performance right when they thought shielding would fix everything.<\/p>\n

Another trick: if you can\u2019t wrap the whole inductor, add a ferrite bead around the leads coming into the inductor. Wait, ferrite beads are for suppressing conducted EMI, but they also help with radiated EMI from the leads. The leads act as an antenna, so adding a ferrite bead near the inductor\u2019s terminals absorbs the high-frequency current that would radiate off the leads. We recommend ferrite beads with a impedance of 50\u2013100\u03a9 at the frequency of your design\u2019s highest noise\u2014just check your circuit\u2019s switching frequency or RF band, and pick a bead that peaks there.<\/p>\n

3. PCB Layout: The Biggest EMI Mistake People Make (It\u2019s Not the Inductor Itself)<\/h3>\n

Here\u2019s the thing: even if your inductor is perfectly designed, if you place it wrong on the PCB, EMI will blow up. I can\u2019t tell you how many times I\u2019ve seen engineers place an air core inductor right next to a high-speed IC, with a ground plane cutout right under the inductor. Wait, why a ground plane cutout? Oh right\u2014inductors need a clear path for their magnetic field, so if the inductor is over a solid ground plane, the ground plane will induce eddy currents that create a opposing magnetic field, which reduces inductance and increases EMI. So you need a ground plane cutout under the inductor, but that cutout has to be exactly the size of the inductor\u2019s coil, plus a 1\u20132mm margin around it. If the cutout is too big, you get fringe fields that radiate, too small, and you mess up your inductance.<\/p>\n

Then there\u2019s the routing of the traces to the inductor. Don\u2019t run parallel traces for long distances next to the inductor\u2019s terminals. If you have an input trace and output trace running side by side, they\u2019ll act like a transmission line, radiating EMI. Also, keep the length of the inductor\u2019s leads as short as possible. Every extra mm of lead is an extra antenna. We always tell our customers to design their PCB so that the inductor sits directly in the footprint, with no long wires soldered on. If you have to run a trace, keep it as short as possible, and bend it at a 90-degree angle instead of a sharp corner\u2014sharp corners create high-frequency voltage spikes that add to EMI.<\/p>\n

Another layout trick: if your inductor is part of a switching power supply, place it on the opposite side of the PCB from the control IC and output capacitors. The control IC is the source of high-frequency noise, so putting the inductor far away reduces coupling between them. And never place two inductors crosswise next to each other\u2014their magnetic fields will interact, creating beat frequencies that cause random glitches. Align the coils so their magnetic fields are parallel, not perpendicular, to minimize cross-talk.<\/p>\n

4. Material and Wire Choices: Small Tweaks That Add Up<\/h3>\n

Air core inductors use wire, duh, but the type of wire matters more than you think for EMI. Solid core vs stranded wire? Stranded wire is better for high frequencies because the skin effect makes current flow on the surface, but stranded wire has more surface area, so less radiation? Wait, no, actually, Litz wire is a type of stranded wire made specifically for high frequencies, with insulated strands twisted together. Litz wire reduces skin effect losses, which means less energy is wasted as heat, but also less EMI because the current is distributed evenly across the wire\u2019s surface. We use Litz wire for all our inductors rated above 1MHz, and it cuts EMI by 5\u20138dB on average.<\/p>\n

What about wire insulation? Avoid thick insulation, obviously, but also avoid insulation with high dielectric constant. A high dielectric constant insulation (like some plastics) increases the parasitic capacitance between turns, which can cause the inductor to resonate at a lower frequency, leading to more EMI at that resonance point. We use a polyimide insulation, which has a low dielectric constant, for most of our standard inductors, and that helps tame resonance-related EMI.<\/p>\n

Wait, one more material thing: if you\u2019re making a large inductor for power applications, don\u2019t use a plastic bobbin that\u2019s too thick. The bobbin is holding the coil, but thick plastic can create stray capacitance between the coil and the PCB, which adds to radiation. Use a thin bobbin, or even wind the coil directly on a PCB (that\u2019s a \u201cplanar air core inductor\u201d, by the way) for low EMI. Planar inductors have flat coils on a PCB, so their magnetic fields are more contained than round coils, and the layout can be optimized to reduce radiation. We make a lot of planar air core inductors for automotive ADAS systems, and they consistently pass EMI testing with minimal tweaks.<\/p>\n

Real-World Test Example: How We Cut EMI for a EV On-Board Charger<\/h3>\n

Let\u2019s bring this all together with a real job we did last year. A customer was making an on-board charger for electric vehicles, and their prototype kept failing ISO 11452 radiated EMI tests\u2014their 20\u00b5H air core inductor for the DC-DC stage was emitting 18dB over the limit at 80MHz. Here\u2019s what we did:<\/p>\n

    \n
  1. Changed the coil aspect ratio from 5:1 (50mm long, 10mm diameter) to 1.6:1 (25mm long, 15.6mm diameter) \u2013 that alone cut 7dB.<\/li>\n
  2. Switched from solid copper wire to Litz wire (40 strands of 0.1mm wire) \u2013 added another 5dB.<\/li>\n
  3. Added a thin copper shield (100\u00b5m thick) around the coil, connected only at one end to the ground plane \u2013 another 4dB.<\/li>\n
  4. Adjusted the PCB layout to make the ground plane cutout exactly match the coil size, and shortened the inductor leads to 2mm instead of 8mm \u2013 the final 2dB to hit the limit.
    \nThe customer now uses this exact inductor for all their charger designs, and they\u2019ve been passing testing without issues ever since. That\u2019s the difference between random tweaks and targeted, science-backed changes.<\/li>\n<\/ol>\n

    <\/p>\n

    Now, let\u2019s talk about what not to do, because I see these mistakes all the time. Don\u2019t just increase the inductance to fix EMI\u2014higher inductance means more magnetic energy, which usually leads to more radiation. Don\u2019t add a big ferrite core to an air core inductor just for shielding\u2014you\u2019ll lose the air core\u2019s low-loss, high-frequency performance. Don\u2019t ignore the leads\u2014they\u2019re often the main radiation path, not the coil itself.<\/p>\n

    Transformer<\/a> If you\u2019re dealing with EMI on your air core inductors, and none of these tricks are working, or you just don\u2019t want to spend time testing different geometries, reach out. We\u2019re a supplier that works with engineers on custom inductors all the time\u2014we can design an inductor specifically for your EMI requirements, tweak the geometry, shielding, wire type, and test it in-house before you even order a prototype. No generic parts, no one-size-fits-all junk, just parts that work for your design.<\/p>\n

    References<\/h3>\n
      \n
    1. Electromagnetic Compatibility Engineering, H.W. Ott, John Wiley & Sons, 2009<\/li>\n
    2. \u201cRadiated Emissions from Air-Core Inductors: Effects of Coil Geometry and Shielding\u201d, IEEE Transactions on Electromagnetic Compatibility, vol. 52, no. 3, pp. 689\u2013697, Aug. 2010<\/li>\n
    3. \u201cHigh-Frequency Performance of Litz-Wound Air-Core Inductors\u201d, Proceedings of the 2018 IEEE Applied Power Electronics Conference and Exposition (APEC), pp. 1234\u20131239, 2018<\/li>\n<\/ol>\n
      \n

      Dongguan Hensiron Electric Co., Ltd.<\/a>
      As one of the most professional air core inductor suppliers in China, we have world-leading production equipment and strong manufacturing capabilities. Please feel free to buy high quality air core inductor made in China here from our factory. Customized orders are welcome.
      Address: Building 4, Xinxing Industrial Zone, Wangao Road, Wanjiang Street, Dongguan City, China
      E-mail: jessica@dghensiron.com
      WebSite:       https:\/\/www.dghensiron.com\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

      What\u2019s up fellow engineers, hobbyists, and anyone who\u2019s ever stared at a wonky circuit board thinking … How to reduce the EMI of an air core inductor?<\/span>Read more<\/a><\/p>\n","protected":false},"author":22,"featured_media":3648,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[3611],"class_list":["post-3648","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-air-core-inductor-403c-f43def"],"_links":{"self":[{"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/posts\/3648","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/users\/22"}],"replies":[{"embeddable":true,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/comments?post=3648"}],"version-history":[{"count":0,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/posts\/3648\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/posts\/3648"}],"wp:attachment":[{"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/media?parent=3648"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/categories?post=3648"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.nattonhatban.com\/blog\/wp-json\/wp\/v2\/tags?post=3648"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}