4

Introduction

I've recently entered the prosumer/pro audio market by buying JBL LSR305 studio monitors and noticed that the tweeters produce constant high-pitched noise (most people apparently call it ‘hiss’) that can be heard even from two metres away. See this spectrogram for a graphical representation. The noise is present with and without input and practically regardless of gain.

Disappointed, I've been researching the issue and it appears that the noise is intrinsic to all Class-D amplifiers and, more worryingly, to all Class-D based monitors. I've learned the basics of audio amplification and the differences between the classes here and found some very helpful information here and here.

Intriguingly, some people claim to either not hear the noise or hear it only when they put their ears right to the tweeters! Could be hearing sensitivity, could be something else.

Question

Naturally, my objective is to remove or at least attenuate the noise, but the question is more general and related to the amplifier itself, hence hopefully relevant.

  1. Since some people claim to not hear the noise, what variables aside from individual differences in hearing could cause it?

Andy aka wrote: “Class D amps are extremely susceptible to hum on the power feeds to the amplifier […] and any AC voltage superimposed on these rails (ripple aka hum) gets transferred to the speaker.”

  1. Could it be that my electrical supply is too noisy? (I live in a flat; 230V/50Hz.) If yes, can I buy/build something to filter it?

Some people claim that using a balanced connection solves the issue, others disagree. It's present even with no input, though.

  1. Can the nature of the input (balanced vs. unbalanced) be related to the issue?

Thank you!

Community
  • 1
  • 3
Harold Cavendish
  • 149
  • 1
  • 4
  • Suffice to say that this most likely not caused by your mains. Buy different monitors w/o class D built-in amp. Or better, buy un-amped "monitors" and build/buy a linear amp with a linear power supply. Unless you plan to demolish the walls of your flat (with SPL), chances are you won't notice the difference on your electricity bill. And I'm guessing you didn't make spectrogram yourself, or you wouldn't be asking this kind of question. – the gods from engineering Oct 18 '15 at 22:39
  • As for balanced connection... if it makes the noise without any signal source/cable plugged into it, then that surely won't fix it. Otherwise, how good is your source in terms of SNR? Can you listen to it (your source) with anything else? – the gods from engineering Oct 18 '15 at 22:46
  • @RespawnedFluff First off, I'm not sure what ‘kind of a question’ do you mean but I did make the spectrogram myself (not sure how it's relevant, though). Second, you provide no useful explanation that would help me understand the issue. The SNR of my source is 115 dB and yes, I can listen to it with something else… without the noise. Finally, I'm obviously trying to understand the origin of the noise and why some people hear it and others don't, not merely solve my issue by buying something else. Thanks. – Harold Cavendish Oct 18 '15 at 23:54
  • The origin of the noise is the amplifier itself then. A class D amplifier is switching circuit, quite capable of generating high-frequency noise precisely of the kind you've posted in that spectrogram. – the gods from engineering Oct 18 '15 at 23:59
  • Even when powered from a battery which gives "clean" DC, you an get noise on a class D amp, e.g. https://youtu.be/STyUxXCIE5o?t=225 – the gods from engineering Oct 19 '15 at 00:01
  • @RespawnedFluff So are the people who hear it only when they put their ears right in front of the tweeters practically deaf or is there perhaps some other variable that could render this feature more pronounced? Perhaps ethernet over power (EOP)? Thanks. – Harold Cavendish Oct 19 '15 at 00:10
  • There is substantial variation between people in sensitivity to high frequency sounds [and noise]. I'll see if I can a paper on that, but basically the older you are, the worse your high-frequency hearing gets. But it's very late where I am and tomorrow is a Monday. – the gods from engineering Oct 19 '15 at 00:12
  • @RespawnedFluff No need to prove that, I'm rather familiar with physiology. I was just trying to find out more about things that I know little about, i.e. electrical components. Thanks for your help. – Harold Cavendish Oct 19 '15 at 00:14
  • @RespawnedFluff Do you have a moment, please? I've bought another pair of monitors with Class AB amplifiers and they hiss as well… It's the same kind of hiss, maybe not as strong. Is this normal too or could it be related to improper grounding (I'm asking because my USB ports and onboard audio outputs conduct noise affected by CPU and GPU utilisation, mouse movement, keyboard backlight, etc. even through an external DAC and it's been suggested to me that it may be a grounding-related issue)? – Harold Cavendish Oct 21 '15 at 21:56
  • Could be normal (or not) depending on their SNR spec. Please post another question with those, with details. If I got this right, the noise varies with PC activity? If that's so, you do have a problem with (PC power supply switching) noise being transmitted to the amp/monitors. You need to have at least one good common ground to avoid that, but it may not be enough, depending on how much noise is being generated and the noise rejection spec of the amp. – the gods from engineering Oct 21 '15 at 22:10
  • For example, if I put a 10K/10K resistive voltage divisor out of an E-MU 1212m, and I pick the signal 10K lifted from the ground (i.e. in parallel to the high-side resistor), I get noise in my amp, even though this card has an amazing 120db SNR. But if I pick the signal with a common ground (i.e. in parallel with the low-side resistor): no noise whatsoever. And this even with a lowly TDA2030A amp made on a stripboard (which needs a resistive divisor because it has high input sensitivity, in case you wonder how I came up doing that). – the gods from engineering Oct 21 '15 at 22:18
  • Also, having ferrite chokes at both ends of an unbalanced audio cable helps somewhat when you take the signal out of a noisy PC source (I'm not talking about the E-MU here). – the gods from engineering Oct 21 '15 at 22:28
  • @RespawnedFluff Okay, thanks. Another question is coming up. Some of the noise does indeed vary with PC activity (more like a buzz with changing frequency and amplitude), some doesn't (the ‘Class-D hiss’ in the JBLs, the Pioneer amplifier from the video that you referred to, and now the Class-AB KRKs). I've already tried a single ferrite bead and the effect is minimal, sadly. – Harold Cavendish Oct 21 '15 at 23:43
  • If the monitor has any sort of signal source or even a cable connected, then you don't know for sure where the noise is coming from. If it's connected to a computer then the source could very well (in whole or part) be the computer. Try your experiment again with pins 2 and 3 of the input XLR shorted together. – little_birdie Jun 24 '16 at 20:25
  • 1
    It's worth pointing out that Class-D amplifiers do not inherently have some sort of unavoidable hiss. While all amplifiers will have some sort of noise (see Johnson noise), well designed Class D amplifiers can have very low noise floors. There are many products with class D amplifiers that have almost imperceptible hiss at full gain. It's not so much that you need "non class D" amps, you just need better-designed monitors. – uint128_t Jan 29 '17 at 19:54
  • 1
    The best example I can give is comparing Mackie SRM450s with QSC K12s. I used to run the Mackies as tops supporting orchestra; during concerts, the hiss of the Mackies was audible at the back of the venue when no one was playing. Switched to K12s and the hiss was only audible when I was ~10ft away. Both are Class-D, the K12s simply have a much lower noise floor. – uint128_t Jan 29 '17 at 19:59
  • @uint128_t That's very interesting, thanks for sharing your experience. I wonder why would a company like JBL with a huge R&D budget and presumably lots of experienced engineers design their speakers without this in mind. I'm assuming that material requirements to ensure there's no hiss can't be significant—or are they? – Harold Cavendish Feb 03 '17 at 09:51

0 Answers0