Hi,
Actually, I'm not scared of people ripping off anything, as such. I jsut disapprove of the practice and refuse to teach such students.
Plus, it will be SO MUCH MORE SATISFYING once you work it out.
You think of non-ideal transformers but in an ideal environment.
For anyone not using Class A/AB amplifier based mains regeneration the environment is very much non-ideal.
So think about how to make a transformer deal better with the non-ideal environment. Equally, think about how the the supplied Circuit (rectifier etc.) interacts with the transformer and the non-ideal environment.
It actually is not really difficult, in fact it needs nothing that I did not learn in EE101. What it does need though is a "system" view (shall we call it electronic holism?) instead of the reductionism that is so prevalent a malaise in modern science...
No.
Low Leakage (or "Ultra Low Leakage") does not mean "no leakage". If you take a MC capable phono stage with -152dBV theoretical Ein you will find that 2mm mild steel can shielded ultra low leakage transformers followed by 3mm mild steel partitions still allow a lot of local flux leakage and still require extreme care of layout.
What such a layer of Cow Metal does, it cuts the field 30dB or so. Of course, the field is ultra-low, but did we not talk about 120dB dynamic range systems elsewhere?
Where-ever it will prevent capacitive coupling. BTW, it is also possible to design mains transformers with very high leakage inductance to reduce RF coupling plus some tricks to maximise common mode rejection.
Knowing how to wind a very good audio transformer does help to wind a good mains transformer, just remember, with a mains transformer you want many of the opposite quality of an audio coupling transformer.
Ciao T
Haven't yet thought of one, so are you going to spill the beans oh trafo sifu? Or still running scared from people ripping off your ideas?
Actually, I'm not scared of people ripping off anything, as such. I jsut disapprove of the practice and refuse to teach such students.
Plus, it will be SO MUCH MORE SATISFYING once you work it out.
You think of non-ideal transformers but in an ideal environment.
For anyone not using Class A/AB amplifier based mains regeneration the environment is very much non-ideal.
So think about how to make a transformer deal better with the non-ideal environment. Equally, think about how the the supplied Circuit (rectifier etc.) interacts with the transformer and the non-ideal environment.
It actually is not really difficult, in fact it needs nothing that I did not learn in EE101. What it does need though is a "system" view (shall we call it electronic holism?) instead of the reductionism that is so prevalent a malaise in modern science...
No.
Low Leakage (or "Ultra Low Leakage") does not mean "no leakage". If you take a MC capable phono stage with -152dBV theoretical Ein you will find that 2mm mild steel can shielded ultra low leakage transformers followed by 3mm mild steel partitions still allow a lot of local flux leakage and still require extreme care of layout.
What such a layer of Cow Metal does, it cuts the field 30dB or so. Of course, the field is ultra-low, but did we not talk about 120dB dynamic range systems elsewhere?
Where-ever it will prevent capacitive coupling. BTW, it is also possible to design mains transformers with very high leakage inductance to reduce RF coupling plus some tricks to maximise common mode rejection.
Knowing how to wind a very good audio transformer does help to wind a good mains transformer, just remember, with a mains transformer you want many of the opposite quality of an audio coupling transformer.
Ciao T
You seem to forget that most people are simply trying to "do-it-yourself" something, and are not going to run copying and selling your ideas. Look at the name of this forum: DIY Audio.
This is the third or fourth time in this thread you suggested you know some extremely good way of doing something - but when asked for details, you refuse to explain anything.
With this kind of attitude you might as well not say anything at all.
I, for one, am not even trying to learn all the underlying physics - it's enough for me that something works, and it doesn't bother me what is the equation behind it. Thus, I have no chance of working things out on my own, like you're telling me to. 95% of what I know about audio, and electronics in general, I've learned by listening to others, not by working it out myself.
And I'm not going to try and cash in on my knowledge or ideas. I'm sharing everything I find, so that whoever is in a similar pursuit as I am, can use this thread as a guide, or at least resource. That is the idea behind internet forums as I understand it. Saying "I know how to do it, but I'm not telling you" is not helping anyone.
Sorry if I'm being harsh, but that's the way I see it.
Hi,
In germany we know kinds of "full wave" rectification.
One is the graetz bridge, which needs four diodes and only one winding (or a centertapped winding to make a +/- Supply).
The other is the traditional full wave rectifier that requires just two diodes but requires two windings (or a centertapped winding).
As you have double bi-filar center-tapped windings (basically 22V-0-22V + 22V-0V-22V if you separate out the bifilar windings) you can do what I suggest.
Two TO-247 double schottky diodes, each on a center-tapped winding, but electrically separate.
Single "big" reservoir capacitor for each rail.
Small local cap at the star ground, the electrically separate supplies are joined at the star ground.
Of course, even better, also go dual mono with separate transformers per channel (which I think you already do).
All I know is that this kind of circuit sounds better than more modern and more economical circuits.
There are some ways (shall we call them trick's) to reduce the issues in the more common circuitry, but forgive me for keeping them for my "putting bread, butter and meat on the table" projects.
What I suggest above is what I use in my "cost no real object" designs, like the AMR AM-77. if you already commission (and pay) for custom transformers, why not do it at the best possible level?
Ciao T
In germany we know kinds of "full wave" rectification.
One is the graetz bridge, which needs four diodes and only one winding (or a centertapped winding to make a +/- Supply).
The other is the traditional full wave rectifier that requires just two diodes but requires two windings (or a centertapped winding).
As you have double bi-filar center-tapped windings (basically 22V-0-22V + 22V-0V-22V if you separate out the bifilar windings) you can do what I suggest.
Two TO-247 double schottky diodes, each on a center-tapped winding, but electrically separate.
Single "big" reservoir capacitor for each rail.
Small local cap at the star ground, the electrically separate supplies are joined at the star ground.
Of course, even better, also go dual mono with separate transformers per channel (which I think you already do).
All I know is that this kind of circuit sounds better than more modern and more economical circuits.
There are some ways (shall we call them trick's) to reduce the issues in the more common circuitry, but forgive me for keeping them for my "putting bread, butter and meat on the table" projects.
What I suggest above is what I use in my "cost no real object" designs, like the AMR AM-77. if you already commission (and pay) for custom transformers, why not do it at the best possible level?
Ciao T
Hi,
What you seem to not know, but of which I am keenly aware - most of the more tractable and well documented designs here are routinely ripped off by Chinese individuals and companies and resold as their own invention.
Luckily most of my designs are so extreme, they are hard to copy commercially (or indeed even implemented commercially by myself).
Actually, things that are easily implemented DIY I tend to give enough details to do so, for anyone who understands the basics. Often in fact I share the most extreme solutions freely.
Much of the rest is of a really "fundamental understanding" nature.
Had I just kept copying without thinking myself I'd still not understand.
Of course I could give you explicit winding data for a transformer that would be "the bees knee". But, as it so happens i do not have the time to design you a personal transformer.
Instead I am telling you how to get (to my knowledge) the best out of the transformer that you have and have paid for and I am telling abraxalito to be less rash in his judgements and to thing more, as he basically questioned the underlying principles behind the design of transformer you ordered and which my previous writings may or may not have influenced, but which I agree with.
Past that, there is wikipedia, all major rectifying circuits are covered in nauseating detail.
Okay, sure. From now on I will not tell you anything. Your wish is my command.
Which is why I give you personally very practical advise, as my free time allows.
Others seem more interested in understanding the principles. Rather telling such people what to do, I am trying to make them think for themselves, first, because they have already proven that they want to and can, second, because ten people looking at the same problem come up with ten ways to solve it.
I could just teach them my way. I could just make them my students amd make myself a Sifu, Guru. I absolutely refuse. With people who want to figure things out, I feel helping them (and at times challenging them) helps them to develop their individual views and ways and the world is richer for it (I hope).
My, my. I know this your thread and I apologise introducing off-topic elements. But I am actually not particulary telling YOU to work it out. We just have several conversations going in this thread and some of these started outside of it and get dragged in here and some are at right angles.
I personally would still suggest that understanding may benefit you in more ways than you realise, but I have accepted your refusal to do so earlier.
So again, I tell you what I know and can communicate on doing the best with what you have. Take it or leave it. Your call.
I know you will not.
Funny story. I published some J-Fet Circuits here for a phonostage (under a different ID from this one, which we may call nom du plume).
The text quite clearly spelled out they where theoretical designs, not build, but should work, if build correctly and using a certain grade of Fet.
I received several very bad auto translated "chinglish" emails via the diyaudio system asking me for help, because the "person" could not make the circuit work. I tried to help.
At the time I was visiting a chinese factory that was making products a friends/business partner of mine was importing.
I found I stood next to the guy who had been e-mailing me.
The factory had taken my purely theoretical design from this board, made 1000 pcs, used fake parts that did not meet any specification I had written about (they where not even within a mile of spec), found nothing worked and afterwards had abused my good-will further by trying to get me to fix their all-across-the-board screwups remotely, for free of course.
So forgive me being somewhat more circumspect since.
Sorry, but that is not what I am saying.
What I am saying is:
"If you want this technology properly you must think about it, understand it or you must pay someone else to do that job for you.
And I'm not actually telling you, I am telling abraxalito, who I know can think this through.
But anyway, you told me not to say anything here, unless I am willing to spend the time to educate everyone or to give a away for free a design everyone can just copy without thinking.
I take it you would like a PCB Layout, preferably so I include the design files? So you can find the lowest lowest bidder, who will not produce the PCB's as I specified anyway. Basically, instead you get a thin, cheap and bad material pcb, that will not produce the correct parasitic capacitances and resistances (which are in my designs usually accounted for) and basically will mean the whole thing does not work.
So, unless I am willing to do it your way, I am asked to shut up. So shut up I will.
Best regards Thorsten
What you seem to not know, but of which I am keenly aware - most of the more tractable and well documented designs here are routinely ripped off by Chinese individuals and companies and resold as their own invention.
Luckily most of my designs are so extreme, they are hard to copy commercially (or indeed even implemented commercially by myself).
Actually, things that are easily implemented DIY I tend to give enough details to do so, for anyone who understands the basics. Often in fact I share the most extreme solutions freely.
Much of the rest is of a really "fundamental understanding" nature.
Had I just kept copying without thinking myself I'd still not understand.
Of course I could give you explicit winding data for a transformer that would be "the bees knee". But, as it so happens i do not have the time to design you a personal transformer.
Instead I am telling you how to get (to my knowledge) the best out of the transformer that you have and have paid for and I am telling abraxalito to be less rash in his judgements and to thing more, as he basically questioned the underlying principles behind the design of transformer you ordered and which my previous writings may or may not have influenced, but which I agree with.
Past that, there is wikipedia, all major rectifying circuits are covered in nauseating detail.
Okay, sure. From now on I will not tell you anything. Your wish is my command.
Which is why I give you personally very practical advise, as my free time allows.
Others seem more interested in understanding the principles. Rather telling such people what to do, I am trying to make them think for themselves, first, because they have already proven that they want to and can, second, because ten people looking at the same problem come up with ten ways to solve it.
I could just teach them my way. I could just make them my students amd make myself a Sifu, Guru. I absolutely refuse. With people who want to figure things out, I feel helping them (and at times challenging them) helps them to develop their individual views and ways and the world is richer for it (I hope).
My, my. I know this your thread and I apologise introducing off-topic elements. But I am actually not particulary telling YOU to work it out. We just have several conversations going in this thread and some of these started outside of it and get dragged in here and some are at right angles.
I personally would still suggest that understanding may benefit you in more ways than you realise, but I have accepted your refusal to do so earlier.
So again, I tell you what I know and can communicate on doing the best with what you have. Take it or leave it. Your call.
I know you will not.
Funny story. I published some J-Fet Circuits here for a phonostage (under a different ID from this one, which we may call nom du plume).
The text quite clearly spelled out they where theoretical designs, not build, but should work, if build correctly and using a certain grade of Fet.
I received several very bad auto translated "chinglish" emails via the diyaudio system asking me for help, because the "person" could not make the circuit work. I tried to help.
At the time I was visiting a chinese factory that was making products a friends/business partner of mine was importing.
I found I stood next to the guy who had been e-mailing me.
The factory had taken my purely theoretical design from this board, made 1000 pcs, used fake parts that did not meet any specification I had written about (they where not even within a mile of spec), found nothing worked and afterwards had abused my good-will further by trying to get me to fix their all-across-the-board screwups remotely, for free of course.
So forgive me being somewhat more circumspect since.
Sorry, but that is not what I am saying.
What I am saying is:
"If you want this technology properly you must think about it, understand it or you must pay someone else to do that job for you.
And I'm not actually telling you, I am telling abraxalito, who I know can think this through.
But anyway, you told me not to say anything here, unless I am willing to spend the time to educate everyone or to give a away for free a design everyone can just copy without thinking.
I take it you would like a PCB Layout, preferably so I include the design files? So you can find the lowest lowest bidder, who will not produce the PCB's as I specified anyway. Basically, instead you get a thin, cheap and bad material pcb, that will not produce the correct parasitic capacitances and resistances (which are in my designs usually accounted for) and basically will mean the whole thing does not work.
So, unless I am willing to do it your way, I am asked to shut up. So shut up I will.
Best regards Thorsten
Steve,
Really?
Interesting!
Split bobbin transformers cancel capacitive coupling through the core?
That news to me. I guess I need to recalibrate that park of test equipment that does say the opposite?
Ciao T
Really?
Interesting!
Split bobbin transformers cancel capacitive coupling through the core?
That news to me. I guess I need to recalibrate that park of test equipment that does say the opposite?
Ciao T
now who's trying to score cheap points?
split bobbin xmfr can have nearly 10x less pri-sec parasitic C compared to toroidial so it is a cheap option - definitely preferable to cheap toroids without a shield on the mains HF noise coupling front
also there is the question of where to connect the electrostatic shield - you have higher parasitic C between it and either winding so it will need to effectively gnd the (higher compared to split bobbin) noise current
split bobbin xmfr can have nearly 10x less pri-sec parasitic C compared to toroidial so it is a cheap option - definitely preferable to cheap toroids without a shield on the mains HF noise coupling front
also there is the question of where to connect the electrostatic shield - you have higher parasitic C between it and either winding so it will need to effectively gnd the (higher compared to split bobbin) noise current
Last edited:
Look, I did not mean to offend, and if I did, I apologise. I appreciate your knowledge and I know that there are people who would abuse your good will, so I understand your reasons for withholding some information.ThorstenL said:
What I was pointing out is that advice like:
is not going to help anyone (and surely not those with limited knowledge of physics) unless you explain the principle behind it. If you don't, then not only you annoy your readers, but also waste your own time writing advice that no one can use.
I'm not saying you should do all the thinking for me. All I'm asking is for a direction which I can research - like when you pointed me at Malcolm Hawkesford's paper about cables, which I'm very grateful for because it then led me to some very interesting discoveries.
You're never going to eliminate it. But as jcx points out, the interwinding capacitance is much lower than that of a toroid without a shield.
And as he also points out, that shield needs to be tied to something. If it's tied to a common ground, well, then you kind of defeat the isolation you're trying to provide as those noise currents can couple through the grounding system.
A split bobbin transformer provides low interwinding capacitance without using an electrostatic shield that can cause as many problems as it cures.
From Bruce Gabrielson and Mark Reimold of Sachs/Freeman:
The single Faraday shield controls all manner of evils which could be attributed to the electric coupling of noise through a transformer. However, the problem with a single shield arises when it is bonded to the ground of either the primary or secondary side of the transformer. The inclusion of a Faraday shield between the primary and secondary windings eliminates inter-capacitance, but it also establishes two new capacitances between the shield and both windings. These two capacitances allow high frequency currents to flow in the grounding systems of both the primary and secondary. Bonding the transformer shield to either the primary or secondary ground establishes current paths for high frequency noise in the reference conductor of the circuit to be isolated. The particular choice of ground for connection of the shield only provides selection of the quieter of the primary and secondary circuits. In many applications, this current path defeats any isolating effect, which a transformer might provide.
se
Last edited:
I must have not been clear - my transformer secondaries are sec.1: 0-22V and sec.2: 0-22V (i.e. separate windings, not centre tapped).
What I initially thought you suggested was:
An externally hosted image should be here but it was not working when we last tested it.
But it probably wouldn't make much sense...
What about if you connect it to actual earth? I mean a thick wire dug directly into earth. Would that not solve everything nicely?
Last edited:
Sorry if I'm being thick, but I have two red bifilar cables and two blue bifilar cables, each pair labeled as 0-22V.
As far as I can comprehend, by separating the wires in each cable, I get four pairs of 0-22V (two of which are in phase with each other and out of phase with the other two). Where does the centre tap come in?...
Your quote says "The particular choice of ground for connection of the shield only provides selection of the quieter of the primary and secondary circuits.". Hence my thought of earth as the quietest possible "circuit". Why does it not solve anything?
Connect two red cables in series (0 with 22v). The connection point is your center tap, same for blue cables.
It doesn't solve anything because a wire stuck in the dirt doesn't have any relevance here. The ground isn't the magical one-way sink for noise that some seem to believe it is.
se
Perhaps a greater understanding of psychology (your own in particular) would help here. All disapproval is rooted in fear. The person who fears nothing, disapproves of nothing.
I do agree, as a teacher myself, but suppose that there really is nothing to work out here? Suppose its engineering nonsense? I'd be inclined to pose a question which elicits the lateral shift in thinking my student needs.
I most certainly do not think in that way. But are you here suggesting that I should?
I took apart an amplifier (actually two, they were monoblocks) that had such a scheme - it was a Mark Levinson. I never could figure out why anyone would go to such an expense. Still I cannot even now some 15 years later.
Well let me think - I've seen a fair few rectifier circuits where there are series resistors in use - either in the AC side (fairly rare) or after rectification (more common). Having a higher resistance secondary might save those components. The non-ideal environment here is mains distortion or RF hash or both?
Its clear to me you don't have a 'system view' when you disapprove of people copying your stuff. Nor when you claim the world is 'non-ideal' (just to offer two examples) so I'm somewhat doubtful that you'll bring a holistic paradigm to audio. But I do digress
I do though agree that western science does itself a disservice by so uncritically embracing reductionism. That's one reason I like audio and am not an establishment scientist.You learned that in EE101 too?
Sure, I aim for 120dB dynamic range as a ballpark in my own designs. But this isn't so much about dynamic range as the absolute level of the leakage flux - in an MC stage of course where the maximum signal level is 60dB or more below the maximum signal in a digital system then its substantially more important. I can see the value of the metal screen in an MC pre power supply, just can't yet see its value in a power amp unless the transformer is fairly close to the circuitry.
Well that was my question. Where in those types would it?
Go on then, what's the detail?
Perhaps though he does know this but, like me, just doesn't care. Perhaps like me he doesn't see anything wrong in this. You do, but so far have not provided a single reasonable explanation for why. When I asked about this on another thread, you became somewhat defensive and threw up various smokescreens but didn't disclose reasons which made any kind of sense. So I conclude its just your irrational fear which leads you to hold back in sharing information. It may not be fear of being ripped off, it could just as easily be fear of being shown to be wrong in some respects which you've long held close to your heart.
This is true, but here people don't just want to copy - they really want to understand. So they not only want suggestions, they'd like reasons behind those suggestions. At least that's my perception.
I don't believe that's what uncle_leon wants. From how I read him, he'd really like to know how to specify his own transformer, to understand what's important to sound quality and what's not. Better to teach a man how to fish...
I find a lot of ThorstenL's comments make me chuckle for their ironic content. Here's a case in point - I haven't made any rash judgements but I have raised some questions. Its interesting that he doesn't notice the difference. Also I teach my own students to think less - thinking tends to be a way to try to head off mistakes before they happen. Its more efficient to make the mistake and learn from it, in my own experience.
So yes, Thorsten, I did indeed question (not rashly judge) specifying only a thicker secondary. You've said there are reasons for this, but to date that's only a claim, you've avoided giving them. So in some sense we're all the poorer - you lost the chance to explain something fairly important, and we lost the chance to learn.
Its true you do give very practical advice. But that's not all you give and I don't think uncle_leon is complaining about the practical advice. Rather he's saying its pointless for you to say 'I have methods for dealing with this, but I'm not going to reveal them'. I agree with him 100% on that point.
It seems to me you have a fundamental misunderstanding of the nature of the Sifu or Guru. Its not a title, its not something that one claims for oneself, rather it is bestowed upon you by your students. No Sifu makes someone their student, the Sifu rather seeks out devoted students and invites them to join him (or her). The invitation though is merely a recognition of what's already happening in the relationship between the student and the master. As the old Zen saying goes 'When the student is ready, the teacher will appear'. So your refusal to 'make myself a Sifu' is totally a red herring.
Yes, but ask yourself - does saying 'I know this but I'm not sharing' help do any of this?
At this point, I'd be inclined to say that as it was a publicly shown circuit, the person ask his questions in public on the forum. That way, others could be helped not to make the same mistakes. Doing it in private was a loss for the wider community.
Strange definition of 'good will' you have. Doing things - that is performing services - for free just makes people value your service less. Your business sense here is rather weird - why would you not charge for your service of fixing the design? Now if you felt responsible for publishing a design which you knew didn't work, I might begin to understand...
Clearly though here they chose the wrong FET, so the service you could provide for them would be in knowing how to obtain the right part. Such a service is definitely chargeable.
That fails to work as a justification for withholding information, as I've pointed out. It just shows you have poor business sense when dealing with the Chinese. Now that's where I could provide you with one of my services (never for free of course)
But abraxalito already knows that he must understand, not merely copy. Yet if I was your student and you really were my Sifu, you'd know not to suppress your student's curiosity by calling it a 'rash judgement' and you'd also know when you were avoiding a question, as you've done here.
Defensive nonsense. What's eating you?
Perhaps I can help make this a little clearer. The interference we'd like to dispose of in this case is high frequency. A spike in the ground will need to have a wire to connect to it, it will necessarily be fairly long as very few people have bare earth in their listening room. A long wire is inductive so won't look like earth to the high frequencies inside the amplifier (where it matters) even if the earth itself is well watered for low impedance.
Does this help?
Thanks abraxalito, your understanding of my intentions is 100% correct, and I agree with your comments above.
Although I also feel I need to emphasize again that I respect Thorsten, and I appreciate his advice.
Could it still bring any potential improvement to the shielding quality?
Although I also feel I need to emphasize again that I respect Thorsten, and I appreciate his advice.
Thanks for explaining that. But what about the case where your listening room is on the ground floor and you have wooden floors and no basement. So that the length of the wire before it makes contact with literal earth could be as short as 1-2m. Which is, incidentally, the case with my listening room
Could it still bring any potential improvement to the shielding quality?Thanks Daniel! It seems so obvious now!...
As far as I can see, you have respected him in your writing. He certainly seems to have taken offense, but you haven't given any that I can see.
For this we can do a quick sum. A 2m length of wire could have an inductance of 2uH. At 1MHz, such a wire will look like an impedance of j12.5ohms. I'd say probably it isn't worth it, but then I don't know yet which RF frequencies are the worst offenders in degrading audio quality. My initial guess is in the band 1MHz - 30MHz - your earth wire will look too high impedance in that bandwidth. To deal with RF in this band, I'd say its better to look to ferrites, which absorb the energy and turn it into heat. Also fit a mains input filter which will contain a common-mode inductor.Thanks for explaining that. But what about the case where your listening room is on the ground floor and you have wooden floors and no basement. So that the length of the wire before it makes contact with literal earth could be as short as 1-2m. Which is, incidentally, the case with my listening room
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.