{"id":188129,"date":"2024-03-11T16:30:12","date_gmt":"2024-03-11T15:30:12","guid":{"rendered":"https:\/\/www.hwcooling.net\/?p=188129\/"},"modified":"2025-11-24T20:55:07","modified_gmt":"2025-11-24T19:55:07","slug":"which-combo-whines-the-least-12-psus-3-graphics-cards","status":"publish","type":"post","link":"https:\/\/www.hwcooling.net\/en\/which-combo-whines-the-least-12-psus-3-graphics-cards\/","title":{"rendered":"Which combo whines the least? 12 PSUs, 3 graphics cards…"},"content":{"rendered":"

“After replacing the PSU, the video card coils stopped whining”. An extremely common statement and yet it always remains on a subjective level, which is a pity. But that is also why we will now look at things in an exact, numerical way. We do this by testing several graphics cards, where for each, noise levels of VRM coils are monitored with each ATX (3.0) power supply used. So how is it with the “whining”? When is it stronger and when is it weaker?<\/p>\n

One tip that often appears in internet discussions on how to solve the louder coil whine in graphics cards is to use a different PSU. For some this may sound like a good idea and something they are willing to experiment with, for others it may sound like something that doesn’t make much sense. We were in the latter camp, or (let me drop the authorial plural) me personally. Rather than speculating about something in a “what could it be” way, it’s better to test things out and then draw some conclusions. For starters, though, let’s indulge in a bit of theory.<\/p>\n

Why a PSU can have an impact<\/h2>\n

First, let’s explain why it makes sense to discuss the dependence of coil (graphics card) whine intensity on a used PSU (ATX). Before that, let’s briefly talk about why the coil makes the noise in the first place. The coil itself is basically a wire wrapped around a metal core. You can’t see it on current components because of the use of shielding (that rectangular box), but it’s exactly like that inside. This coil has a voltage applied to it, which creates vibrations. These, because of the design of the coils as such, are the source of the noise. How this noise will sound depends, besides the electrical properties, also on the degree of vibration damping by glue, stabilization.<\/p>\n

\"\"<\/a>
Historical illustration of an old motherboard with Socket 7. You can’t see coil threads from the outside on current components…<\/figcaption><\/figure>\n

If the coils were perfectly coated with it, there would be no vibration, no sound. However, to achieve such a state in practice would be extremely expensive, unprofitable, and that is why coils more or less “whine”. In fact, rather than “whistling”, it is a kind of “whirring”. It is in fact a mix of different frequencies of sound, not one distinctive one (like when you blow a whistle, for example).<\/p>\n

\"\"<\/a>
… it’s not possible without shielding these days. It’s around coils to eliminate mutual electromagnetic interference<\/figcaption><\/figure>\n

Coils are noisier or quieter depending on whether they are “better” or “worse” at hitting resonant frequencies at a given job. When they do hit them, they can be quite noisy.<\/p>\n

For a more clear idea, we can probably note that coils have some “natural” resonant frequency that arises at perfectly clean voltages. But this (clean, smoothed voltage) is only a theory, and practice may be such that there is always more or less electrical noise. The latter is defined by a ripple of some amplitude (peak to peak in millivolts) and then a time period, or frequency. And we assume that it is the latter (the electrical noise frequency) that can interfere with and affect the “base frequency” of the coils. Negatively, but also positively. The nature of the electrical noise itself does not say whether the coils will be noisier or quieter due to its presence. Equally “aggressive” noise can lead to both results. It is only a question of whether, in interaction with it, the coils will be closer to or further from the resonant frequencies at which the loudest noise is achieved.<\/p>\n

\"\"<\/a>
Visualization of voltage noise by an oscilloscope. The waveform of the 12-volt branch is yellow<\/figcaption><\/figure>\n

Now, on to why, in theory, none of what we wrote above would carry weight. There could be no dependence of coil whine on the power supply used, as long as there is perfect voltage smoothing at the level of the graphics card VRM ahead of the coils (on the capacitors), or smoothing to a degree that does not affect the coil vibration in any way.<\/p>\n

\"\"<\/a><\/p>\n

However, such a situation can also be considered as purely theoretical and in practice there will be qualitatively different voltages on the coils, which may have an impact on the noise of the coils. Whether it’s dramatic or merely symbolic (or something in between) will only become clear in tests. One must prepare for these (tests) properly, so as not to end up measuring something different than what one wants. But we will get to that in the text of the next chapter with the testing methodology.<\/p>\n