John Curl's Blowtorch preamplifier part II

[john curl]

Sooner or later, eddy current losses and their effects will be better understood by my 'critics' here. One only has to measure the Q of the coils with frequency to understand it almost completely, and of course, that is what I was specifically looking for in my quest in the JAES for a graph that was clearly defined and explicit. Alas, I did not find such a graph, just its cousins and uncles, and only an 'expert' could interpret them to get the specific information necessary. All the 'cat calls', cartoons, or any other impugning of the work that I once did at Ampex, so many decades ago, are for naught.

[jneutron]

Originally Posted by jcx

my disagreement is with the "teleological" slant, apparent disregard of basic Linear (EM) systems theory fundamentals

No problemo..we differ in opinion, ok..

Originally Posted by jcx

in a Linear model resistor noise is Not signal depedant

I assume you speak of thermal and not shot.

Originally Posted by jcx

- eddy currents are just the consequence of conductive materials/resistances coupled by mutual inductances

Your statement is for the most part accurate, however a more generalized explanation is required to explain incident planar waves for example, and cylindrical wires as well..but still, not a bad one..

Eddy currents are caused by a conductors response to a time varying magnetic field.

Originally Posted by jcx

so "eddy current noise" would be the thermal noise of the resistances so coupled

A "current", regardless of what or where it is, will produce shot noise.

The real concern here, is...will the topology of the system allow the eddy current shot noise to couple back? This is a basic conundrum not unlike the coupling between a coax shield and it's center conductor..The center conductor is incapable of knowing if there is any shield current, but the shield will be in the center conductor's field lines..

Originally Posted by jcx

arguing from HP/Aligent's model that is so simplified that it can be silk screened on the instrument panel and only requires the 2 values of a single complex number is beside the point

First, I am not arguing. I merely pointed out that a well designed instrument who's function is to determine effective Ls/Rs, considers losses caused by eddy currents to have a significant impact on the measured series resistance.

Second, the manual is from 1988...Was Aligent even around then??

Originally Posted by jcx

the "eddy current noise" has to "be there" all of the time and should appear as frequency dependant noise at the transformer terminals - with or without applied signal - as a result of Linear EM theory

If I push DC into the transformer there are no eddy currents, therefore no "eddy current noise". If I push hf AC into the transformer, there will be eddy currents. If one postulates the existance of "eddy current noise", one must accept the fact that if there are no eddy currents, no noise...and if there are eddy currents, there will be noise.

Cheers, John

[jneutron]

Originally Posted by gpapag

"Eddy Current Losses in Transformer Windings and Circuit Wiring"
http://focus.ti.com/lit/ml/slup197/slup197.pdf

Regards
George

Thanks for the link. It is a somewhat rudimentary paper, but as steve said, it does not focus on noise of the eddy currents..

Better papers on eddy loss and xfmr designs can be had by Dr. Sullivan of Dartmouth..

Originally Posted by Steve Eddy

What we're talking about is audio transformers, which are using wires on the order of 42 gauge. With such small diameter wire, skin and proximity effect are effectively non existent at audio frequencies.

Agreed.

Originally Posted by Steve Eddy

And again, it's not the windings that are at issue, but rather the alleged added noise due to eddy current losses in the core.se

To me, eddy generated noise is not "alleged". However, you are correct in stating that an audible impact on devices being discussed is indeed alleged. I would have liked to see specific papers and research on the topic.

Originally Posted by DF96

With regard to eddy current losses, one way to think of these is to add an extra winding to the transformer then put a resistor across the winding.

This setup guarantees that any noise generated within the secondary will couple to the primary via flux.

Eddy currents do not couple back in this direct fashion, so I would not use this method. (good thinking however, I also considered this test but after thinking a while, understood the limitations.)

What is needed to evaluate this is:
1. Wind 4 C cores with pickup windings. One core solid iron, one 1 mm laminations, one 1 mil laminations, and one comprised of ferrite.
2. Excite the core using a rotating neodymium magnet. The drive must be low noise.
3. Examine the output waveforms for indications of noise amplitude consistent with frequency.

If noise increases in cores which have higher eddy current losses but not in low loss cores, then the hypothesis of eddy current based shot noise is consistent.

If noise does not increase despite the varying core constructs, the hypothesis may not be consistent.

Originally Posted by john curl

Sooner or later, eddy current losses and their effects will be better understood by my 'critics' here.

Oddly enough, nobody is criticizing you. It has been pointed out that you made allegations of eddy based noise, but you have not provided any content to back that assertion up. I happen to concur with you in the assertion, but that and a quarter....

Hey, it was asserted that I am neglecting basic linearity..from jcx ""apparent disregard of basic Linear (EM) systems theory fundamentals""
...that wasn't an attack, it was part of a discussion..he pointed out very clearly what he believes I neglected.

(now, the eddy current monsters...that's an attack..)

Originally Posted by john curl

One only has to measure the Q of the coils with frequency to understand it almost completely,

Actually, no. However, that measure will indeed show the differences in the cores, so this measure, while proportional to the alleged effect, does nothing to explain the effect. Q does not measure the propensity to noise, just dissipative losses.

Cheers, John

[john curl]

Moving on to completion of phono playback, 2 stage design:
While the relatively simple circuit put here by Joachim, is ideal for diy people, it is NOT SOTA in sophistication that we have been making commercially for the last 30 years, yet it still can be an excellent example of quality phono design, IF one final factor is kept in mind. Please use HIGHEST QUALITY coupling caps. More sophisticated designs, REMOVE the need for these caps, and we add servos, instead. Don't let people tell you that you can use just about ANYTHING. Enough said.

[john curl]

I think that this completes the example of 2 stage phono stage. This will get you 90% there, like a Honda or Toyota. Mercedes is something else, but very few here, need it.
When I am designing Constellation Audio products, I am designing at Mercedes, Bentley, etc. level, and small things like input transformer differences, make a difference for my customers. This is why I reached out for opinions on input transformers. Thanks everyone for your input.

[scott wurcer]

Originally Posted by jneutron

Excite the core using a rotating neodymium magnet. The drive must be low noise.

Funny, yesterday I was thinking of the old slide the supermagnet down the copper pipe experiment but you could never get the stiction noise low enough. I would try dropping the magnet in a vaccum.


As soon as you talk about loss (in this case heat) noise is implied. I had a discussion here about shot vs Nyquist noise in this case and it was not obvious to anyone. Take a noiseless 1000v battery an complete the circuit with a 1M resistor and the noise in the 1mA current is < sqrt(2qI) .

Another thought experiment that might relate to what jcx is talking about. Take a coil placed over a conductor. I would postulate that thermal motion of carriers in both would exchange energy at equlibrium (equipartion theorem) exactly like the eddy current. This does happen with capacitors, I have measured it i.e. if one plate has sheet resistance it will induce charge in the other plate (JFET channel resistance induces gate current noise at high f).

[jneutron]

Originally Posted by scott wurcer

Funny, yesterday I was thinking of the old slide the supermagnet down the copper pipe experiment but you could never get the stiction noise low enough. (edit:you added vacuum)

I love that experiment...so cool..edit: hmm, I wonder if the pipe were encased in a bismuth lining..

Originally Posted by scott wurcer

As soon as you talk about loss (in this case heat) noise is implied.

Agreed. The biggest problem I see, is how to decouple the increased heatloss (noise) as a result of proximity effect current density increase in the driving conductor--- from any increase in eddy current loss noise within laminations.

Originally Posted by scott wurcer

I had a discussion here about shot vs Nyquist noise is this case and it was not obvious to anyone.

And that somehow surprises you???

Cheers, John

[DF96]

I think the name "eddy current noise" is causing confusion. It is not noise caused by the presence of eddy currents, but noise caused by the possibility of eddy currents. The simple fact that there is an energy dissipation mechanism present is sufficient to create thermal noise from that mechanism. The mechanism does not have to be dissipating energy from a signal. The mechanism just has to exist. That is all. Just like a resistor; its thermal noise does not depend on the voltage across it, just its temperature.

Current does not produce shot noise. It is uncorrelated discrete charge carriers which create shot noise. Almost all currents use highly correlated charge carriers, so no shot noise. The exception is when charge carriers have to cross a potential barrier, which shields them from each other and reduces correlation. Even in a normal valve shot noise is heavily suppressed by the space charge.

If eddy current noise is present, then it is thermal noise not shot noise. If eddy currents are a loss mechanism (which they are) then they will produce thermal noise. Sorry, the possibility of eddy currents will produce thermal noise.

As Q is a measure of dissipative loss, it will correlate with thermal noise. In fact, this correlation is itself evidence that the noise is thermal in origin even if we have not pinned down the exact mechanism.

If John Curl measures lower Q and hears more noise, then that suggests that the noise is thermal in origin. If he also finds that thinner laminations give higher Q and less noise, then it is quite likely that the mechanism is eddy current losses.

[scott wurcer]

Originally Posted by DF96

Current does not produce shot noise. It is uncorrelated discrete charge carriers which create shot noise. Almost all currents use highly correlated charge carriers, so no shot noise. The exception is when charge carriers have to cross a potential barrier, which shields them from each other and reduces correlation. Even in a normal valve shot noise is heavily suppressed by the space charge.

Carver Mead argues to the equivalence of the two (just to add confusion).
This is a good read.

http://www.rle.mit.edu/avbs/publicat...journal_16.pdf

[DF96]

Yes, what is called shot noise sometimes turns out to have a thermal origin (e.g. the hot cathode in a valve). I have long suspected that valve 'shot noise' is actually thermal noise arising from the space charge. There is genuine shot noise too, in other circumstances (e.g. noise diode). No time to read that article now, but I will look later.