Mixed up measuring and audio, sorry. With regards to audio: eventually you will probably want your complete chain to have balanced IO's.
I have not looked at the details, but the abstract says "... suppresses radio and electro-magnetic frequencies from contaminating the final audio program signal." There is no such thing as electro-magnetic frequencies, so suppressing these is moot 🙄. Suppressing radio frequencies can be nice, but won't help much with typical ground loops, which are at mains frequency (typically 50 or 60 Hz), which are well below radio.
I can imagine a device that detects 50 Hz ground current in the uA or mA range, and if such current detected it disconnects the ground wire. But questions arise: for how long? And what about electrical safety in the lack of permanent ground connection?
About electro-magnetic fields and frequencies:
https://www.bfs.de/EN/topics/emf/introduction/introduction_node.html
https://www.bfs.de/EN/topics/emf/introduction/introduction_node.html
So, there was a typo and you choose to deny the science on that basis? Suppose it said "electromagnetic energy frequencies" instead; would you still deny the science?
How do you know they are mostly 50 or 60Hz? Because you can see AC line harmonics on an audio FFT, but you can't see radio frequency spurs? And if you can't easily see it as a PSS spur on an audio FFT then it doesn't exist, or it can't have any audible effect?
I wouldn't bother you or others about this stuff except it isn't good for newbies and lurkers to read these threads and come across improperly dismissive, incorrect and or misleading claims which naive readers may then believe are true. Of course, innocent mistakes are understandable, but we should at least try to keep working towards improving a correct understanding of science. Too many people today have lost confidence in science to the point they feel whatever they want to be believe is just as good as what scientists say. That's how society degenerates to the point of believing things like flat earth theories. People read it on some internet forum and then they believe it.
Anyway, for people interested in keeping up with the science, it turns out AC lines can act like pretty good radio wave frequency receiver antennas. And plenty has been published on the types of problems it can cause in audio systems, with some of the work of Bill Whitlock being just one example. The problem is that some of the effects can be hard to directly measure. That's why pulsed RF testing is done on new opamp designs; its a simplified way to check for certain RF sensitivities.
Last edited:
Sounds a lot like a GFI (Ground Fault Interrupter), except a GFI disconnects the power instead of the ground.
It's a patent. Very technical and precise language. No typos, and no science in there.
(I learned this the hard way when I did a patent myself.)
Mains AC is 50 Hz or 60 Hz in most (all?) countries on this planet, so that's what you get in ground loops. If you see higher order harmonics they are very likely due to rectifier noise that couples into the audio signal. However, that's not related to ground loops.
We all know about rectifier noise in FFTs. How do you know there isn't also RF getting into the system via AC line antenna effect, with audible effects such as Bill Whitlock described? IME, most people have no idea if they have a problem with that or not. IIRC, according to Whitlock you can only find out if you have/had a "veil" on your sound once you get rid of it. Again, my experience is similar to his on that point.
Regarding the science, do you accept or deny the notion its practically possible to have an effect as suggested by: "... electro-magnetic (energy/wave) frequencies... contaminating the final audio program signal." Why were you dismissive about it?
Last edited:
Electromagnetic frequencies is like electromagnetic strawberries, i.e., nonsense. But we're going off topic. Let's stop here.
With that type of logic, one could argue that in medicine urinary frequency is like urinary strawberries. In that case the doctor might want to refer the patient for some kind of therapy.
Its actually pretty simple, frequency is how often something happens. Cyclic energy transfer between electric and magnetic energy storage happens in various contexts, and it happens with some frequency. Sometimes there are multiple frequencies involved. For one example, if there are two radio stations broadcasting AM radio then there are multiple electromagnetic frequencies propagating through the atmosphere. For each AM station there is an electromagnetic carrier frequency, a modulator frequency, and two sideband frequencies. They all happen to be electromagnetic phenomena which exist at particular frequencies.
I think everyone here should understand the above. The nonsense is to argue nothing like that exists in the physical world.
With that, I will stop. And agree its time to get back on topic.
Its actually pretty simple, frequency is how often something happens. Cyclic energy transfer between electric and magnetic energy storage happens in various contexts, and it happens with some frequency. Sometimes there are multiple frequencies involved. For one example, if there are two radio stations broadcasting AM radio then there are multiple electromagnetic frequencies propagating through the atmosphere. For each AM station there is an electromagnetic carrier frequency, a modulator frequency, and two sideband frequencies. They all happen to be electromagnetic phenomena which exist at particular frequencies.
I think everyone here should understand the above. The nonsense is to argue nothing like that exists in the physical world.
With that, I will stop. And agree its time to get back on topic.
While RF interference entering in thru the AC cord is possible, it rarely happens.
More likely is RF interference being generated by the audio component and being transmitted by the AC cord.
More likely is RF interference being generated by the audio component and being transmitted by the AC cord.
It ingress or egress, it can go either way. But, yes, class-D and or possibly SMPS, is more likely an emitter.
Regarding being in the vicinity of transmitters, it can sometimes be plenty to have neighbor who shares the same power company distribution transformer as your house to be running a vacuum cleaner, or similar. Elevators in apartment buildings that have them are often bad electrical noise emitters.
Just using good common mode chokes on each AC plug at the wall or power strip can have clear and obvious corrective effect on ground loops and or other AC line noise sources which have been producing an unknown "veil" or "grainy" sound in the reproduction system (the words "Veil" and "grainy" having been taken from a Bill Whitlock AC line noise AES presentation). Adding a little more AC line filtering at the AC cord power source can in some cases offer additional improvement. Depends on the filter design and the particular noise environment. However, just failing to filter or choke one or two AC line cables can ruin the whole effort. IOW, it has to be done right to offer the most benefit.
EDIT: It is only in the most severe cases that a radio transmitter of some type is clearly heard coming out of the audio reproduction system. However, a severe problem is the only type many people realize may be AC line related. Again, the words Bill Whitlock used are veiled and or grainy. IIRC he also said in a recorded panel discussion in response to a question about it that the only way you know you have or had a veiled or grainy sound is after you fix it. IME, he is right. "A previously unknown veil was removed," is exactly how it happens sometimes. Same as "the wife comes out of the kitchen and asks what did you change?" Sometimes it happen just like that. Its where such stories come from in the first place, from real events.
Regarding being in the vicinity of transmitters, it can sometimes be plenty to have neighbor who shares the same power company distribution transformer as your house to be running a vacuum cleaner, or similar. Elevators in apartment buildings that have them are often bad electrical noise emitters.
Just using good common mode chokes on each AC plug at the wall or power strip can have clear and obvious corrective effect on ground loops and or other AC line noise sources which have been producing an unknown "veil" or "grainy" sound in the reproduction system (the words "Veil" and "grainy" having been taken from a Bill Whitlock AC line noise AES presentation). Adding a little more AC line filtering at the AC cord power source can in some cases offer additional improvement. Depends on the filter design and the particular noise environment. However, just failing to filter or choke one or two AC line cables can ruin the whole effort. IOW, it has to be done right to offer the most benefit.
EDIT: It is only in the most severe cases that a radio transmitter of some type is clearly heard coming out of the audio reproduction system. However, a severe problem is the only type many people realize may be AC line related. Again, the words Bill Whitlock used are veiled and or grainy. IIRC he also said in a recorded panel discussion in response to a question about it that the only way you know you have or had a veiled or grainy sound is after you fix it. IME, he is right. "A previously unknown veil was removed," is exactly how it happens sometimes. Same as "the wife comes out of the kitchen and asks what did you change?" Sometimes it happen just like that. Its where such stories come from in the first place, from real events.
Last edited:
I have been reading Bill Whitlock papers since 1995. I don't recall him writing some of those things.
Your recollection is correct. Here are two examples found through use of a search engine.
Bill Whitlock Generic Seminar Notes
Bill Whitlock Magazine Article
It seems like a bad translation of EMI or electro-magnetic interferance. That is a catch-all phrase that usually means stuff 30 MHz and below. That stuff can couple magnetically (hum fields) and electrostatically from inadequate shielding. However The spectrum of EMI has no limits actually. Its an unfortunate use of words.
In the US, good patent practice is to define your terms in the patent to prevent this kind of confusion. Unfortunately some terms can mean very different things when used in a different context. e.g.: For example, "bank" can refer to a financial institution or the edge of a river.
A link to the 2012 Bill Whitlock seminar:
https://indianaaes.org/wp-content/uploads/2012/09/indy-aes-2012-seminar-w-notes-v1-0.pdf
I can't find a link to his 2019 seminar.
https://indianaaes.org/wp-content/uploads/2012/09/indy-aes-2012-seminar-w-notes-v1-0.pdf
I can't find a link to his 2019 seminar.
Looks like he took out the reference to "veiled" and "grainy." Wonder why. Pushback from engineers demanding "proof" because they don't see it in an FFT?
If so, there can be two or three reasons I can think of offhand for it not clearly showing up well in an FFT (some possible reasons may be mutually exclusive).
1. Its not a PSS (periodic steady state) signal, so typical FFT analysis assumptions don't hold (because FFTs measure average bin frequency correlation; thus non-PSS signal energy is spread across multiple or many bins).
2. Its lost because of discarded phase information in a spectral magnitude view.
3. Its included in the noise skirts seen at the base of spectral lines.
In any case, I know from my own experience that Whitlock was right when he talked about "veiled" or "grainy" sound, and I also know from my own experience he was likely to get pushback from engineers who don't really understand what FFTs do show well versus what FFTs don't show well.
If so, there can be two or three reasons I can think of offhand for it not clearly showing up well in an FFT (some possible reasons may be mutually exclusive).
1. Its not a PSS (periodic steady state) signal, so typical FFT analysis assumptions don't hold (because FFTs measure average bin frequency correlation; thus non-PSS signal energy is spread across multiple or many bins).
2. Its lost because of discarded phase information in a spectral magnitude view.
3. Its included in the noise skirts seen at the base of spectral lines.
In any case, I know from my own experience that Whitlock was right when he talked about "veiled" or "grainy" sound, and I also know from my own experience he was likely to get pushback from engineers who don't really understand what FFTs do show well versus what FFTs don't show well.
Last edited:
I dunno... Veiled and grainy is right there in his article from 2019, updated in 2020. I'd guess that if he wanted to dump veiled and grainy from the discussion, he had two chances right there.
2019 Whitlock Article
BTW, most modern audio analysis software packages, like REW and Multitone to name just two free examples out of many, have the provision to capture and save peak values. If you take enough samples over a long period of time, you can usually see some of these elusive issues. This is true not only for audio, but in communications systems not related to audio - digital, analog, and mixed signals. You can catch events that lead to degraded transmission that way. Averaging, ahh, averages out random events into virtual nothingness. That doesn't mean that they aren't real or having an effect, like dropping data packets, for example.
But, I have very rarely ever seen spectral captures of the electrical output of consumer audio systems. That is, at the power amplifier or loudspeaker terminals. (I'd love to have somebody point to some examples so that I can see what I've missed.) Usually, only components of the system are measured in the best kind of Heisenberg isolation boxes that the measurer can manage, to get the best results. That can be really accurate, but doesn't show the entire picture. By a long shot.
2019 Whitlock Article
BTW, most modern audio analysis software packages, like REW and Multitone to name just two free examples out of many, have the provision to capture and save peak values. If you take enough samples over a long period of time, you can usually see some of these elusive issues. This is true not only for audio, but in communications systems not related to audio - digital, analog, and mixed signals. You can catch events that lead to degraded transmission that way. Averaging, ahh, averages out random events into virtual nothingness. That doesn't mean that they aren't real or having an effect, like dropping data packets, for example.
But, I have very rarely ever seen spectral captures of the electrical output of consumer audio systems. That is, at the power amplifier or loudspeaker terminals. (I'd love to have somebody point to some examples so that I can see what I've missed.) Usually, only components of the system are measured in the best kind of Heisenberg isolation boxes that the measurer can manage, to get the best results. That can be really accurate, but doesn't show the entire picture. By a long shot.