See PMA below
See PMA below
Bingo.The rate of change of current is max at zero crossiing, you state it accurately.
I believe that would be correct. I suspect a big spike at the cusps of course. Difficult to separate a reading from inductive spikes though..
jn
jn
You have all my respect when it comes to EM.
But this generation of harmonics didn’t cling to me from the start.
I have some days now reading back Heaviside’s Electromagnetic Theory and I didn’t find any supportive evidence. The closest to the subject is para 191, pages 344-349 (*) of Vol. I (I haven’t looked at his Electrical Papers yet).
The explanation you give here (in bold letters)
is central to your reasoning.
I want to ask:
1. Is what you suggest a result of experiments or of theory?
2. In case of experiments Is it possible that harmonics are caused by (a) heat diffusion velocity effects in the conductors and/or (b) attraction - repulsion forces causing minute distance modulation btn conductors?
My understanding is that if electr. conductivity is very high (and magn. permeability is 1), this time varying resistance will be synchrounous and with no delay in respect with the excitation, thus it will affect current only in amplitude - and this a little- and not generate new frequencies.
George
>Edit (*) At page 347 talking of surface current flow (skin effect in our times), he states that resistance per unit area of surface varies directly as the square root of the product of resistivity, inductactivity and frequency.
You have all my respect when it comes to EM.
But this generation of harmonics didn’t cling to me from the start.
I have some days now reading back Heaviside’s Electromagnetic Theory and I didn’t find any supportive evidence. The closest to the subject is para 191, pages 344-349 (*) of Vol. I (I haven’t looked at his Electrical Papers yet).
The explanation you give here (in bold letters)
is central to your reasoning.
I want to ask:
1. Is what you suggest a result of experiments or of theory?
2. In case of experiments Is it possible that harmonics are caused by (a) heat diffusion velocity effects in the conductors and/or (b) attraction - repulsion forces causing minute distance modulation btn conductors?
My understanding is that if electr. conductivity is very high (and magn. permeability is 1), this time varying resistance will be synchrounous and with no delay in respect with the excitation, thus it will affect current only in amplitude - and this a little- and not generate new frequencies.
George
>Edit (*) At page 347 talking of surface current flow (skin effect in our times), he states that resistance per unit area of surface varies directly as the square root of the product of resistivity, inductactivity and frequency.
Last edited:
the resistance doesn't change, the (partial, spatially varying) mutual inductance with the proximity current causes emf (also spatially varying) that appears to redistribute the current through the sheet of the physically distributed resistance
the appearance of the current "being shoved to one side" only happens with identical waveforms in the wire and the sheet resistance
the coupling between the physically distinct parts of the same current loop causes a change in AC impedance - since the mutual inductance, physical extent of the R are continuous quantities the frequency dependence is a fractional power series in "s"
proximity effect is bigger in typical loudspeaker cables than the "pure" skin effect - yet we don't attribute any cable nonlinear distortion to the effect there
nonlinear products of the magnitude of the proximity effect impedance rise would wipe out Mb DSL modem QAM signal over thousands of feet of legacy voice grade twisted pair - why did Scott's customers want better than -80 dBc op amps to drive those cables if percent 2nd harmonic were generated by proximity effect?
the appearance of the current "being shoved to one side" only happens with identical waveforms in the wire and the sheet resistance
the coupling between the physically distinct parts of the same current loop causes a change in AC impedance - since the mutual inductance, physical extent of the R are continuous quantities the frequency dependence is a fractional power series in "s"
proximity effect is bigger in typical loudspeaker cables than the "pure" skin effect - yet we don't attribute any cable nonlinear distortion to the effect there
nonlinear products of the magnitude of the proximity effect impedance rise would wipe out Mb DSL modem QAM signal over thousands of feet of legacy voice grade twisted pair - why did Scott's customers want better than -80 dBc op amps to drive those cables if percent 2nd harmonic were generated by proximity effect?
Last edited:
If it is rate of change of current, as jn asserts, rather than frequency, as the textbooks say, then the penetration of current would vary through the cycle. This would mean that any RF or microwave system would generate second harmonic whenever RF encountered a conductor, as the current density would be varying at twice the frequency. I am not aware of this phenomenon ever being reported.
I am baffled why jn continues to assert something which is simply false. The maths is quite clear: the geometric distribution of current depends on frequency alone. Where a sharp pulse is concerned then the individual frequency components (e.g. f, 3f, 5f etc. for a square wave) each have their individual exponential decay into the conductor. This means of course that a sharper pulse may penetrate less but this is because of its higher frequency components not its greater amplitude.
In the case of the resistor test the input was a single frequency so there can be no debate: one frequency means one exponential geometric decay throughout each cycle.
I am getting bored with this. There is a point beyond which it is not worth repeatedly asserting known EM theory to someone who rejects it.
I am baffled why jn continues to assert something which is simply false. The maths is quite clear: the geometric distribution of current depends on frequency alone. Where a sharp pulse is concerned then the individual frequency components (e.g. f, 3f, 5f etc. for a square wave) each have their individual exponential decay into the conductor. This means of course that a sharper pulse may penetrate less but this is because of its higher frequency components not its greater amplitude.
In the case of the resistor test the input was a single frequency so there can be no debate: one frequency means one exponential geometric decay throughout each cycle.
I am getting bored with this. There is a point beyond which it is not worth repeatedly asserting known EM theory to someone who rejects it.
for sure it can be frustrating - but also a opportunity for all to improve their explanatory skills, even their understanding
so far this particular back and forth appears quite the model for what I would like to see here more often in technical discussions/disputes
much better than personal attacks, "black bag" rhetorical twisting of every word, appeal to inscrutable, conveniently unavailable "authorities"...
so far this particular back and forth appears quite the model for what I would like to see here more often in technical discussions/disputes
much better than personal attacks, "black bag" rhetorical twisting of every word, appeal to inscrutable, conveniently unavailable "authorities"...
well where do you run your wire, traces, mount your R , other components when you don't have inches of Al between them and power xfmr, AB output device current paths?
pay any attention to loop area?
I think big currents near steel parts has already been mentioned - but are there other effects with nominally linear conductors, resistive materials "in proximity"?
pay any attention to loop area?
I think big currents near steel parts has already been mentioned - but are there other effects with nominally linear conductors, resistive materials "in proximity"?
Last edited:
Most certainly there are; the conductors may be essentially linear, but the materials around them, in contact with them have marked non-linear behaviours. These are ripe for future, decent investigation; at a practical level this is where a lot of "the action is", as regards improving sound quality ...
Frank
That the indisputable progress of audio technology since 40 years can be credited to better understanding (and application) of physical laws rather than smelling "hype' components ?
I believe that it is with a compilation of knowledge like this, even if it seems usefulness at first sight, that we can design our stuff in a better way: We do not design a printed board the same way if we have a 'physical or instinctive feeling' of the coil and resistance of a track, and the parasitic caps around, than if we do not ? Don't-you ?
Sorry, but on my side, i am more happy to have learned one or two new things i never asked myself about resistance's behaviors than to know, out of any context, that John Curl prefer "audiophile" ones.
[Edit] My answer doubled JCX's one with the same meaning.
Brad, when I re-emerged from the bush, one of the first things I noted in da 21st century was a comment of yours. I enshrined this in a "pearls of wisdom" document of mine. Here it is somewhat paraphrased ..
My interest is in condensor mikes. I know that 4416s aren't as good as they were ... compared to even 5 yrs ago.
Scott, I would be interested if you have looked at gate leakage during your epic exploration of amps for condensor mikes.
The manufacturer is important. In da 21st century, I favour Fairchild but this is simply cos they publish noise specs for all their stuff.[*] This beach bum doesn't have facilities to check loadsa samples today.
30 yrs ago, my favourite FET for condensor mikes was Mullard / Philips BFW11 made by Welsh virgins. Today, no BFW11 is an audio LN device.
[*] But Scott sez they may just be re-branders.
Christophe, I'm having some fun with your MP3 "memories": upsampled, and of course the file sounded better. Decided to go the whole hog, and resampled to 2 x DSD, DSD128, to see how that turned out ... not bad, a 17Meg file became 1.2Gig in size. And, of course this is far superior, to the ear, to the lowly MP3 that we started with ... seriously!
Deaf Harry, a close relative of Blind Freddy, could pick this, so I'm going to stop mucking about, and record the 2 versions via the "crap" PC speakers, on my low life recorder, a cheap digital camera. Then do some decent, comparative analysis of what's going on, no-one else seems to be doing it, so I might as well be the first one ...
I'll see what turns up, cheers,
Frank
Deaf Harry, a close relative of Blind Freddy, could pick this, so I'm going to stop mucking about, and record the 2 versions via the "crap" PC speakers, on my low life recorder, a cheap digital camera. Then do some decent, comparative analysis of what's going on, no-one else seems to be doing it, so I might as well be the first one ...
I'll see what turns up, cheers,
Frank
I would like to clarify something. I do not use AUDIOPHILE anything, except my own topologies that I created to be the best amps and preamps that I could devise.
Rel caps are sold around the world, to government labs, as well as to audio manufacturers.
Penny and Giles makes military pots as well, that is how I discovered that they made rotary pots, was well as their well known linear pots. I told Mark Levinson about them, and the rest is audio history.
Resista had no AUDIOPHILE label until we discovered that it worked better than many brands.
What I recommend WORKS! It has been proven to work by a number of INDEPENDENT audio designers, not just me. Other similar products may work, as well, or maybe not as well. You choose, what works, or what is in your junkbox?
Rel caps are sold around the world, to government labs, as well as to audio manufacturers.
Penny and Giles makes military pots as well, that is how I discovered that they made rotary pots, was well as their well known linear pots. I told Mark Levinson about them, and the rest is audio history.
Resista had no AUDIOPHILE label until we discovered that it worked better than many brands.
What I recommend WORKS! It has been proven to work by a number of INDEPENDENT audio designers, not just me. Other similar products may work, as well, or maybe not as well. You choose, what works, or what is in your junkbox?
John, please, do not take my previous remarks the wrong way. I yet said i had no doubt (who have ?) about your technical knowledge or experience.
The things i pointed is *you never care to justify* your technical assertions by technical argues, measurements and even your listening impressions. And do not answer when asked; like about the different signatures of pots you had found.
More than that, in one message, you said some wired pot was the only one good enough for you, and some time later, you recommended ALPS.
You seem to believe that "used by John Curl" is enough, and you are soooo definitive !
As you do not say where and why, how and why could-we use your recommendations ?
I'm suspicious too because all the stuff you recommend is *Always* the most expensive one. My experience proved it is far to be always true, and, as all poor guys, i'm mostly interested with quality/price ratio..
About my personal efforts, and to answer your question, do-you believe it would be possible to had been promoted as the responsible of the R&d department of electro-acoustic of the most important audio manufacturer of my country, at this time and at the age of 25, without both technical background and a real passion ? i build my first tube amp kit at the age of 12
It is so nice from-you. I am very flattered and honored you spend so much time on this, while I doubt those poor tracks, full of drop out, so destroyed and erased by the time can be restored to sound much better than the old yellowed photography they are. But i would be, of course, very interested to listen to your results, if it is possible.
You will be amused to know that those tracks where first saved on Sony minidisk 20 years ago, then recopied via analog in MP3 for the web. I still have the originals, but they are worse than the Minidisk copy. Thanks to digital and optical with time.
I wish I could produce an animation, that would show so much...
jn
jn,
That would help but I think I'm following your basic argument just not quite how it produces a second.
If one accepts the dI/dT dependency and that this effect only increases the crowding resistance then I can understand how for a sinewave this would result in cos modulation from the dI/dT and the absolute value to account for only increasing resistance.
But from there I see this modulation producing steeper slopes around the cross-overs, local minimums at the peaks, all symmetrical between positive and negative going cycles of the sinewave. Why wouldn't this symmetry measure as a dominant 3rd?
Real or not at least for a series feedback divider the effects should cancel out (between the high side and low side resistors).
Thanks
-Antonio
- Status
- Not open for further replies.