{"id":821,"date":"2026-11-20T05:47:02","date_gmt":"2026-11-19T21:47:02","guid":{"rendered":"http:\/\/www.shepherds-life.com\/blog\/?p=821"},"modified":"2026-11-20T05:47:02","modified_gmt":"2026-11-19T21:47:02","slug":"how-to-reduce-noise-in-switching-power-supply-4fb7-60524a","status":"publish","type":"post","link":"http:\/\/www.shepherds-life.com\/blog\/2026\/11\/20\/how-to-reduce-noise-in-switching-power-supply-4fb7-60524a\/","title":{"rendered":"how to reduce noise in switching power supply"},"content":{"rendered":"<p>Hey guys, it\u2019s Mike here, and I\u2019ve spent the last 12 years working as the go-to switching power supply guy for all sorts of teams\u2014from small EV startups scrambling to hit production deadlines to medical device OEMs that can\u2019t afford a single dB of noise messing with their sensitive sensors. Let\u2019s cut to the chase: if you\u2019re here because your switching power supply is feeding you noise headaches\u2014whether that\u2019s weird hum on your audio equipment, data glitches in your embedded systems, or overheating that\u2019s cutting your product lifespan short\u2014this post is exactly what you need. <a href=\"https:\/\/www.szsbrzdy.com\/enclosed-power-supply\/switching-power-supply\/\">Switching Power Supply<\/a><\/p>\n<p><img decoding=\"async\" src=\"https:\/\/www.szsbrzdy.com\/uploads\/202339560\/small\/power-module-12v-1a0b465e8a-2f35-48ed-8b59-66a5455fbf7a.jpg\"><\/p>\n<p>First, let\u2019s get one thing straight: switching power supplies (SMPS) are awesome. They\u2019re way more efficient than old linear supplies, they don\u2019t get too hot under load, and they save us all a ton of energy. But their whole \u201cswitching\u201d design is the root of the noise issue. Those fast, high-voltage on\/off transitions (called dV\/dt and dI\/dt for the nerds in the room) create electromagnetic interference (EMI) that can wreak havoc on nearby circuits. I see this all the time with new customers who come to us with a prototype that works on the bench but fails EMC testing once it\u2019s in a real enclosure. They think, \u201cJust pick a better part number,\u201d but 9 times out of 10, the fix is in the design tweaks, not the component.<\/p>\n<p>Let\u2019s start with the basics I go over in every customer\u2019s initial design review. No matter what kind of SMPS you\u2019re making\u2014AC-DC, DC-DC, buck, boost, whatever\u2014the biggest noise culprits are the loop areas where high currents switch fast. Think of it this way: current flowing in a loop creates a magnetic field. The bigger the loop, the stronger the field, and the more noise you\u2019re radiating. I can\u2019t tell you how many times I\u2019ve watched an engineer lay out a board with the input capacitor, switching MOSFET, output diode (or synchronous rectifier, these days), and inductor all spread out across two sides of the board. That creates a huge, ugly loop that\u2019s basically an antenna for noise.<\/p>\n<p>Here\u2019s the simple fix: shorten that high-current switching loop. Place the input decoupling ceramic capacitor as close to the MOSFET\u2019s drain and source pins as humanly possible. Same for the output side\u2014put the output capacitor right next to the rectifier and inductor. We often see customers using a bulk electrolytic cap for big ripple, but they forget the small ceramic 0.1\u00b5F or 1\u00b5F cap that sits right at the switch node. That tiny cap is like a noise sponge for the fast switching spikes, and it cuts the loop area by 70% or more without breaking the bank. We even sell pre-assembled decoupling kits for our most popular SMPS models to make this easier\u2014engineers love not having to hunt for the right value and voltage rating.<\/p>\n<p>Next up is grounding, which is like the unsung hero of noise reduction. A lot of folks still use that old \u201cstar ground\u201d method where all ground lines meet at one point, but for SMPS, that\u2019s a recipe for noise coupling. Why? Because the high current from the switching loop flows through the ground plane, and if your small signal circuits (like the ones feeding your microcontroller or sensors) share that same ground, the current creates a voltage drop that shows up as noise on the signal lines.<\/p>\n<p>What works way better for SMPS is a split ground plane, but with a solid connection at only one point (called a \u201cstar point\u201d again, but used properly here) near the input and output. Let the power stage (MOSFETs, inductors, rectifiers) have their own ground plane section, and keep the control circuit (PWM IC, feedback resistors, optocouplers) on a separate ground section. The only place these two sections connect is at that one low-impedance point near the input cap. That way, the high switching noise stays in the power ground, and your sensitive control signals stay clean in their own ground. I always tell customers to test this with a scope\u2014hook a probe to the control ground and a probe to the power ground, and you\u2019ll see a huge difference in noise levels once you fix that split plane.<\/p>\n<p>Now, let\u2019s talk about components, because not all parts are made equal when it comes to noise. I see engineers skimp on capacitors all the time, and that\u2019s a mistake. For input decoupling, use a low ESR (equivalent series resistance) ceramic cap\u2014X7R or X5R dielectrics are perfect, stay away from Y5V if you can. ESR causes resistive noise, so a low ESR cap means less heat and less voltage ripple from the switching current. On the output side, if you\u2019re dealing with audio or precision sensors, pair that bulk electrolytic with a small film cap\u2014film caps have really low inductance, so they handle high-frequency noise way better than electrolytics.<\/p>\n<p>Inductors are another big one. A lot of folks use ferrite core inductors, but if you\u2019re dealing with high-frequency switching (like 100kHz and up), a shielded inductor is non-negotiable. Unshielded inductors radiate magnetic noise like crazy\u2014 I\u2019ve had customers switch from an unshielded inductor to a shielded one and cut their radiated EMI by 20dB overnight. We spec shielded inductors in all our high-noise-susceptibility SMPS models, and it\u2019s one of the most requested upgrades from medical and audio customers. Also, watch out for inductor saturation\u2014if the inductor saturates under load, that creates even more noise and can even cause the supply to glitch. Always derate your inductor current by 20-30% to avoid that.<\/p>\n<p>Then there\u2019s the control loop, which might sound boring, but it\u2019s where a lot of noise leaks in from the feedback circuit. The feedback line from the output voltage divider goes all the way back to the PWM IC, and if that line is long or runs next to a noisy switching line, it picks up interference. The fix here is to keep the feedback trace short, and if you have to route it next to a power trace, put a ground plane between them to act as a shield. Also, use a small capacitor across the feedback resistors (like a 1nF or 10nF ceramic) to filter out high-frequency noise on the feedback signal. I see engineers skip this cap all the time, and the result is the PWM IC overreacting to noise spikes, which makes the switching transitions sharper and noisier. It\u2019s a simple fix, but it makes a huge difference.<\/p>\n<p>Wait, let\u2019s not forget about filtering, both on the input and output. Even with all the layout and component tweaks, sometimes you still need a little extra filtering to pass EMC tests. On the AC input for AC-DC supplies, add an EMI filter at the very front of the supply\u2014this is a common-mode choke plus X and Y capacitors. Common-mode chokes block noise that\u2019s present on both AC lines (hot and neutral) relative to ground, which is the biggest source of conducted noise. We include a compact EMI filter in all our industrial-grade AC-DC SMPS, because those guys have to pass strict EMC standards. For DC-DC supplies, a simple LC filter on the output can smooth out any remaining ripple and noise\u2014just pick an inductor value that works with your load current and a cap that matches the required ripple spec.<\/p>\n<p>I also need to mention switching frequency, which is a balance between noise and efficiency. Higher switching frequencies let you use smaller inductors and capacitors, which is great for size and cost, but they also create more high-frequency noise that\u2019s harder to filter. Lower switching frequencies (like 50kHz instead of 500kHz) reduce that high-frequency noise, but they can make the inductor and output bulkier. The sweet spot these days is around 100kHz to 300kHz for most applications\u2014we design our SMPS to let you adjust the switching frequency if needed, which is a feature a lot of customers don\u2019t realize is there until they come to us with a noise problem and we tweak that setting for them.<\/p>\n<p>Let\u2019s talk about a real example, because that\u2019s how I learned this stuff. Last year, we had a customer who was making a portable ultrasound device. Their SMPS was working great on the bench, but when they put it in the plastic enclosure, the ultrasound sensor was picking up noise from the supply, so the images were blurry. They tried changing capacitors, trying different inductors, nothing worked. We went through their layout together, and found that their switching loop was 2 inches long\u2014they\u2019d spread the input cap, MOSFET, and inductor across three different layers of the board. We redid their layout to shrink that loop to 0.3 inches, switched to a shielded inductor, and added that feedback filter cap. The noise dropped by 32dB, and they passed EMC testing on the first try. That\u2019s the stuff that makes this job worth it\u2014solving real problems for real products.<\/p>\n<p>Now, a few common mistakes I see every week, so you don\u2019t have to make them. First, don\u2019t use vias in the high-current switching loop. Every via adds a tiny bit of inductance, and multiple vias in the loop multiply that inductance. We tell customers to keep vias out of the switching path entirely, or if you have to use them, only use one small via per pin. Second, don\u2019t forget about the input and output voltage ratings. If your caps or inductors are rated too low, they\u2019ll fail under load, which creates noise and safety issues. Always derate voltage by 20% minimum. Third, don\u2019t ignore the secondary side if you\u2019re dealing with isolated SMPS. The optocoupler or transformer feedback line is another big noise path\u2014add a small cap across the optocoupler\u2019s output to filter noise there, and make sure the transformer has a shield winding to block common-mode noise between primary and secondary.<\/p>\n<p>At the end of the day, reducing noise in SMPS isn\u2019t rocket science, but it does require paying attention to the tiny details that most engineers skip. It\u2019s about keeping high-current loops tight, separating sensitive circuits from noisy ones, picking the right components, and filtering where needed. If you\u2019re still struggling, don\u2019t guess\u2014reach out to someone who actually works with SMPS every day, like our team here. We don\u2019t just sell parts; we work with customers to fix their noise issues before they go to production, saving them time and money.<\/p>\n<p><img decoding=\"async\" src=\"https:\/\/www.szsbrzdy.com\/uploads\/202339560\/small\/dc-switching-power-supplya7a65373-60a0-4176-a282-a714eb2f4870.jpg\"><\/p>\n<p>If you\u2019re dealing with a noise problem that\u2019s stalling your project, or you want to learn more about how our SMPS can be customized for low-noise applications, hit us up to discuss your needs. We\u2019re here to help you get your product to market without the headache of EMI or noise failures. And for anyone still scrolling, remember: the best noise fix is the one you apply in the layout phase, not after you\u2019re already in EMC testing. Don\u2019t be the guy who has to re-spin a board because they skipped the decoupling cap close to the MOSFET.<\/p>\n<p><a href=\"https:\/\/www.szsbrzdy.com\/enclosed-power-supply\/\">Enclosed Power Supply<\/a> References:<\/p>\n<ol>\n<li>Pressman, Abraham I. Switching Power Supply Design. McGraw-Hill, 2017.<\/li>\n<li>Electromagnetic Compatibility Engineering. Henry W. Ott. John Wiley &amp; Sons, 2011.<\/li>\n<li>Application Note AN-1021: EMI Reduction Techniques for Switching Power Supplies. Linear Technology, 2000.<\/li>\n<\/ol>\n<hr>\n<p><a href=\"https:\/\/www.szsbrzdy.com\/\">Shenzhen Boerze Power Technology Co., Ltd.<\/a><br \/>Shenzhen Boerze Power Technology Co., Ltd. is one of the most professional switching power supply manufacturers and suppliers in China, also supports customized service with low price. We warmly welcome you to wholesale cheap switching power supply made in China here from our factory. If you have any enquiry about free sample, please feel free to email us.<br \/>Address: Shenzhen Longhua Dalang Anhongji Xingyao Industrial Park<br \/>E-mail: szsbrzdy0620@outlook.com<br \/>WebSite: <a href=\"https:\/\/www.szsbrzdy.com\/\">https:\/\/www.szsbrzdy.com\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Hey guys, it\u2019s Mike here, and I\u2019ve spent the last 12 years working as the go-to &hellip; <a title=\"how to reduce noise in switching power supply\" class=\"hm-read-more\" href=\"http:\/\/www.shepherds-life.com\/blog\/2026\/11\/20\/how-to-reduce-noise-in-switching-power-supply-4fb7-60524a\/\"><span class=\"screen-reader-text\">how to reduce noise in switching power supply<\/span>Read more<\/a><\/p>\n","protected":false},"author":12,"featured_media":821,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[784],"class_list":["post-821","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-switching-power-supply-4eb5-608565"],"_links":{"self":[{"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/posts\/821","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/users\/12"}],"replies":[{"embeddable":true,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/comments?post=821"}],"version-history":[{"count":0,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/posts\/821\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/posts\/821"}],"wp:attachment":[{"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/media?parent=821"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/categories?post=821"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.shepherds-life.com\/blog\/wp-json\/wp\/v2\/tags?post=821"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}