Suppose it depends on what is meant by circuit-based isolation. In the DC supply cases I worked on and for the HS PDS paper, I would say ferrites were used for circuit-based power isolation. Still caused problems.
A power filter schematic sketch can be seen at: https://www.diyaudio.com/community/threads/general-purpose-dac-clock-board.413001/post-7692677 This is one of the cases where ferrites caused problems.
A power filter schematic sketch can be seen at: https://www.diyaudio.com/community/threads/general-purpose-dac-clock-board.413001/post-7692677 This is one of the cases where ferrites caused problems.
Last edited:
I reread your post edited after I made my post. The paper you attach defers to the example of ferrite use between a supply and a circuit, presumably on an umbilical.
I wouldn’t equate this with an example of using ferrite after rectification at the beginning of filtration.
I wouldn’t equate this with an example of using ferrite after rectification at the beginning of filtration.
No umbilical for FPGA power on a HS board. Some of that stuff runs at as much as several GHz. Power regulation has to be on the same ground plane.
The point is that some people appear to hear a real effect. Ferrites can produce hysteresis distortion/noise as described by Bruno Putzeys. Its a property in general of steel, iron, ferrites, etc. There are other effects too. The idea that ferrites merely turn RFI/EMI into heat without any other physical effects seems implausible to me. And something has to explain when people hear real differences in sound. Makes sense to check for known physical effects once convinced at least a few people are likely hearing something real.
After all, if you can produce circuits with and without ferrites, at least it opens up the opportunity to do listening tests between the two different circuits. Of course, professionally conducted formal listening tests are expensive and time consuming so they are rarely done at that level.
Also as I have pointed out before, manufacturers tell what they need to in order to get you to try their products. They figure if the products don't work out for you then at least they tried. OTOH, they don't usually feel obligated to tell you what's wrong with their products. Its up to you to do your own testing. And who does testing on whether ferrites are problematic for audio in some ways? Its taken as a given that they are flawless and always a good thing. Ferrite marketing guys deserve a big raise for pulling that off.
The point is that some people appear to hear a real effect. Ferrites can produce hysteresis distortion/noise as described by Bruno Putzeys. Its a property in general of steel, iron, ferrites, etc. There are other effects too. The idea that ferrites merely turn RFI/EMI into heat without any other physical effects seems implausible to me. And something has to explain when people hear real differences in sound. Makes sense to check for known physical effects once convinced at least a few people are likely hearing something real.
After all, if you can produce circuits with and without ferrites, at least it opens up the opportunity to do listening tests between the two different circuits. Of course, professionally conducted formal listening tests are expensive and time consuming so they are rarely done at that level.
Also as I have pointed out before, manufacturers tell what they need to in order to get you to try their products. They figure if the products don't work out for you then at least they tried. OTOH, they don't usually feel obligated to tell you what's wrong with their products. Its up to you to do your own testing. And who does testing on whether ferrites are problematic for audio in some ways? Its taken as a given that they are flawless and always a good thing. Ferrite marketing guys deserve a big raise for pulling that off.
Last edited:
Do you believe ferrite can be used in certain parts of a supply for audio equipment without negative effects?
Yes. Particularly soft ferrites that let go of stored hysteretic energy in small amounts at a time. Hard and highly attenuating ferrites may produce less frequent yet more impulsive bursts of noise. This is quite possible while both types of ferrite noise might look the same on an FFT. That's because the difference in statistical behavior (impulsiveness) of a given type of noise (say, white noise) is not given by its frequency distribution, rather its a function of its phase distribution, which is in turn a function of the physical process that creates noise in the first place.
It seems this entire thread is about two subjects:
1) Pre-conceived bias that leaves people convinced there is a difference when there is actually none
and
2) Differences that DO factually exist that others cannot believe are possible.
Both categories of situation exist. I understand some people want to warn others to watch their wallets when someone comes along saying they have something to sell that will make a difference. But not all such products are snake oil.
With everything, we have to keep an open mind and not tie our insecurities or fear of humiliation into a debate just so we don't suffer the ego deflation of being wrong.
I've been wrong in the past. My brother proved this to me once when we were testing interconnects. IMO they DO make a difference.
Power cables? I can think of many reasons they wouldn't make a difference but also a few reasons they might. Have I tested any? No. So at this stage I must say that I... don't... know. Because I don't know I'm not going to preach to anyone with all the fervor of a religious zealot.
Everyone has the right to make their own mistakes. Life is a learning process. When people spend money, it's their money. Who am I to say they are wrong? If something pleases someone and makes them happy, why would I want to take that away from them? And I am also fond of the philosophy "judge not lest ye be judged." I also recognize that the same book says that you can't save a fool from himself. I let it be.
I'm into this hobby because I enjoy the technical challenge and having to consider a problem with so many variables. It keeps me busy and entertained, because I love working and I love music.
I also love my friends and colleagues and those who share my interest in this subject, and for all of you I can only say have fun and I wish you all low THD and a high signal to noise ratio.
1) Pre-conceived bias that leaves people convinced there is a difference when there is actually none
and
2) Differences that DO factually exist that others cannot believe are possible.
Both categories of situation exist. I understand some people want to warn others to watch their wallets when someone comes along saying they have something to sell that will make a difference. But not all such products are snake oil.
With everything, we have to keep an open mind and not tie our insecurities or fear of humiliation into a debate just so we don't suffer the ego deflation of being wrong.
I've been wrong in the past. My brother proved this to me once when we were testing interconnects. IMO they DO make a difference.
Power cables? I can think of many reasons they wouldn't make a difference but also a few reasons they might. Have I tested any? No. So at this stage I must say that I... don't... know. Because I don't know I'm not going to preach to anyone with all the fervor of a religious zealot.
Everyone has the right to make their own mistakes. Life is a learning process. When people spend money, it's their money. Who am I to say they are wrong? If something pleases someone and makes them happy, why would I want to take that away from them? And I am also fond of the philosophy "judge not lest ye be judged." I also recognize that the same book says that you can't save a fool from himself. I let it be.
I'm into this hobby because I enjoy the technical challenge and having to consider a problem with so many variables. It keeps me busy and entertained, because I love working and I love music.
I also love my friends and colleagues and those who share my interest in this subject, and for all of you I can only say have fun and I wish you all low THD and a high signal to noise ratio.
I think we are in the same situation: hearing stops at 10-11Khz some tinnitus, dip in one ear at 4Khz, but still enjoining music big time.
Cables: never spend much as my goal was to almost eliminate the issue by going fully active by XLR, except for my electrostatics where I put mono amp close to the speaker, never spend any money on power cables or interconnects.
From the paper: "The frequencies involved in EMI range from 30 MHz to 1 GHz for most products."
There is a huge difference between audio at 200Hz and radio at 200MHz. Audio is not high speed digital. Honestly, I'm glad I don't suffer from these problems. Even in your CD, SACD, or blue ray player the frequencies are too low to matter. If they did matter the TV picture would be affected too.
Does he have a more specific name? Google searches on "someone" and "the guy" didn't really narrow it down too well.
Thanks
Incorrect. Clock edge risetimes in digital audio involve frequencies up to and including several hundred MHz. For dac such as some ESS dacs using 100MHz clocks, risetime related frequencies may extend above 1GHz.
Doesn't matter who. If you really want to know for some good reason maybe I would tell you some more by PM.
Yes to read a 44.1KHz - 196KHz signal. That's kilo-hertz. I don't think a human can hear a 1GHz signal or any smearing from a 1GHz signal or distortion from a 1GHz signal. These are cell phone and microwave frequencies and should not be coming out of an audio amplifier, or across the speaker cables, or in any way transmitted through the atmosphere by anyone's loudspeaker. This is a thread about bi-wiring not microwave effects on audio that some unknown person somewhere talked about. Anyway, I think you should take your DAC argument over to a DAC thread.
It's square/rectangular waves at those frequencies, which have lots of harmonics into many MHz - depending on, as Mark said, risetimes, this generates a theoretically infinite bandwidth of harmonics.
Going the other way, there's many stories of a cellphone next to a cheap small "PC speaker" that makes noise just before the phone rings. GHz signals can easily get demodulated/converted to audible DC pulses in base-emitter diode junctions of audio equipment where the RF frequency is allegedly too high to matter.
Going the other way, there's many stories of a cellphone next to a cheap small "PC speaker" that makes noise just before the phone rings. GHz signals can easily get demodulated/converted to audible DC pulses in base-emitter diode junctions of audio equipment where the RF frequency is allegedly too high to matter.
This is way off topic. GHz interference and noise in a DAC design is not relevant to the placebo effect of bi-wiring.
Sorry but I have to go here, If your DAC is leaking noise from the clock into the analog audio it's simply a poor design, and in a device with a noise floor of 120db the noise is imperceptible by 99.999999% of humans during music playback. Remember the 120Db noise floor is there because anything below that is background heat noise and quantum effects, it's always going to be there.
The cell phone noise is low level switching noise as the phone finds a suitable frequency. This radio conversation is picked up by the analog circuitry in the conference room teleconference equipment and played through the conference room loudspeaker. The conferencing equipment is not shielded well and is usually made of mostly analog circuitry. You should try calling around a radar system! It's common for high level signals to block or interfere with low level signals, this is the basis for jamming in warfare. This is NOT what is happening in a DAC. A DAC should have adequate filtering in place to remove the high frequency analog noise from the signal path before the signal is converted from digital to analog and then it should be shielded.
I'm not going to worry about clock noise when there are things that produce effects orders of magnitude greater like capacitors and power supplies.
Sorry but I have to go here, If your DAC is leaking noise from the clock into the analog audio it's simply a poor design, and in a device with a noise floor of 120db the noise is imperceptible by 99.999999% of humans during music playback. Remember the 120Db noise floor is there because anything below that is background heat noise and quantum effects, it's always going to be there.
The cell phone noise is low level switching noise as the phone finds a suitable frequency. This radio conversation is picked up by the analog circuitry in the conference room teleconference equipment and played through the conference room loudspeaker. The conferencing equipment is not shielded well and is usually made of mostly analog circuitry. You should try calling around a radar system! It's common for high level signals to block or interfere with low level signals, this is the basis for jamming in warfare. This is NOT what is happening in a DAC. A DAC should have adequate filtering in place to remove the high frequency analog noise from the signal path before the signal is converted from digital to analog and then it should be shielded.
I'm not going to worry about clock noise when there are things that produce effects orders of magnitude greater like capacitors and power supplies.
DACs were only one example. There is a significant body of literature to read and understand on the topic of RFI/EMI. It includes various ways cables can be involved. Henry Ott's famous book on "Electromagnetic Compatibility Engineering" would be a good start, but its only a start. Advanced treatment of noise, including signal-correlated noise, is less organized into a single reference.
https://www.booksgoat.com/index.php...14U-L0C6Yn6rPE6n4FbidgKgrVEtRufwaAiIBEALw_wcB
Also, some good suggested supplementary reading: https://www.ti.com/lit/an/sboa128a/sboa128a.pdf?ts=1731356532477&ref_url=https%3A%2F%2Fwww.google.com%2F
Some particular sections of possible interest:
1.1 How EMI Enters Systems and Devices
1.2 How EMI Affects Op Amps
Again, there is more to the subject than can be found in one application note. In particular, simple DC offsets result from continuous wave rectification not from demodulation of a modulated EMI source.
https://www.booksgoat.com/index.php...14U-L0C6Yn6rPE6n4FbidgKgrVEtRufwaAiIBEALw_wcB
Also, some good suggested supplementary reading: https://www.ti.com/lit/an/sboa128a/sboa128a.pdf?ts=1731356532477&ref_url=https%3A%2F%2Fwww.google.com%2F
Some particular sections of possible interest:
1.1 How EMI Enters Systems and Devices
1.2 How EMI Affects Op Amps
Again, there is more to the subject than can be found in one application note. In particular, simple DC offsets result from continuous wave rectification not from demodulation of a modulated EMI source.
Last edited:
Of course it matters. If there's no way to know how this person organizes/prioritizes the elements of what they are hearing , there's no way to to get calibrated in a way that allows useful perspective.
You can certainly choose not to share that info, but by not doing so your comment is wasted.
Nonsense. If you want to know about that I could tell you more than you could learn from googling a name. However I am not going to tell you everything as he considers some of it proprietary. Also, some of it you probably wouldn't be able to understand anyway without spending many hours listening along with him.
Basically, its like this: Its not all about distortion. Its not about about wow and flutter. Its about something else entirely, which is in a single term, "sound stage." It has width, depth, and precision. For that to be at its best, everything has to be right. If its not you can hear the flaws. If there is problematic distortion it will interfere with the soundstage. If it has too much wow and flutter it will interfere with the soundstage. It is only necessary to listen to the soundstage because if anything is wrong the soundstage will show it.
The problem for most people is they have no idea what a well reproduced soundstage can sound like. So they have no point of calibration. Usually they haven't even learned how to listen as well as they think they can to a live unamplified acoustical event. As has been discussed in other contexts here the forum, the brain discards what it considers useless sensory information. The sound of a room is something many people have not learned to hear (just clap your hands once, then listen to the echoes and their decays; practice listening to every room in your house). Especially not if if they are trying to understand someone talking at a low volume level in the same room at the same time. First you have to learn to hear the room, then you have to learn to listen to the person talking or playing an instrument without loosing the sound of the room. By default if you concentrate on listening to the instrument, your brain will discard much or all the sensory information of the room sound. Once you learn how to listen to everything at once, then you have to find a system capable of reproducing that so you know what is possible to encode on a well recorded CD. At that point you can know a good system when you hear one. You are calibrated.
Okay. The above is a brief explanation. I am not going into more detail. If you can't get the basic idea from that, then telling you more is unlikely to help.
Last edited:
There's nothing secretive about a public post. But I suggest listening with our own ears.
- https://www.diyaudio.com/community/...rch-preamplifier-part-iii.318975/post-5825281
- https://www.diyaudio.com/community/...rch-preamplifier-part-iii.318975/post-5825281
- Status
- Not open for further replies.