John Curl's Blowtorch preamplifier part II - Page 1141 - diyAudio
Go Back   Home > Forums > Member Areas > The Lounge

The Lounge A place to talk about almost anything but politics and religion.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 16th March 2011, 01:44 PM   #11401
diyAudio Member
 
jneutron's Avatar
 
Join Date: Jan 2004
Location: away
Quote:
Originally Posted by mrfeedback View Post
Audio transformers are wound with hair thin wire - are you saying that the wniding wire exhibits skin effect change in resistance over the audio band ?.
Ditto, are you saying that the core materials change resistance according to frequency also.

Eric.
I assume this was to me...if not, oh well, I'm answering anyway.

An interesting question.

The guage of wire is sufficiently low that normal skin effect won't be a problem. Most consider skinning based on that exponential approximation equation, with some depth associated with the conductor at some frequency.

However, we've been talking about proximity effect also. Proximity effect is the current being forced to one side of the conductor as a result of the overall field present in the coil. For a very small coil with lots and lots of turns, the magnetic field present on the inner turns of the coil is significantly larger that a single wire field would be, so this enhanced field can still force the current to move to one side of the wire.

It would require a lot of computing horsepower to do the analytical computation needed to find the rise in resistance. We've done it here lots using roxio or opera (not sure which), but it takes half a day to do one iteration..

As for core materials changing resistance with frequency, I'm not sure about that one. Core losses will increase if you increase the rate of change of the flux within it, but that is not the same as lowering the core resistivity.

It is far more reasonable to just use a good meter to determine Ls-Rs over the frequency band of interest, and use the effective resistance vs frequency as the metric.

ps...JC, ya shoulda waited..

Cheers, John

Last edited by jneutron; 16th March 2011 at 01:46 PM.
  Reply With Quote
Old 16th March 2011, 01:58 PM   #11402
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Quote:
Originally Posted by jneutron
If there is no primary current, there is no net flux in the core. No net flux, no rate of change of flux. No rate of change of flux, no eddies. No eddies, no eddy losses.

To accept what you are saying, a mechanism which is capable of eddy current dissipation without any current flowing through any of the terminals MUST be present.
Yes, precisely. The mechanism (magnetic induction, resistive losses) is present, whether it is being driven by a signal or not. This is all you need for thermal noise. I know you can't accept this, but it is true. It is not the fact that eddy currents are currently dissipating energy which creates the thermal noise, but that the mechanism is present and the noise source (the core) is at a non-zero temperature and therefore will have thermal fluctuations.

Take a transformer (or a choke). Connect it to an amplifier or meter, so you can hear/measure noise. Just before you do this, measure the DC resistance and the AC resistance at some appropriate frequency/ies. Measure the noise voltage, and heat the core while trying to avoid heating the windings (not easy, but it will take a while for heat to transfer between them so be quick). You should see the noise increase. Quickly repeat the DC and AC resistance measurements. Measure/estimate the core temperature. Measure/estimate the winding temperature (possibly from any rise in DC resistance). Calculate the thermal noise from the given resistances and temperatures (assume that winding resistance is unchanged from the DC value?). Compare with noise measurements.

Not an easy test, but it is the best I could think of to show the effect. The idea is to show that raising the core temperature increases noise, both by raising the corresponding resistance and raising the temperature of that resistance.
  Reply With Quote
Old 16th March 2011, 02:15 PM   #11403
diyAudio Member
 
jneutron's Avatar
 
Join Date: Jan 2004
Location: away
Quote:
Originally Posted by DF96 View Post
Yes, precisely. The mechanism (magnetic induction, resistive losses) is present, whether it is being driven by a signal or not.
You are basically saying that there will be eddy losses despite there being no magnetic flux rate of change present.

You will have to provide a mechanism to explain that.

Quote:
Originally Posted by DF96 View Post
This is all you need for thermal noise. I know you can't accept this, but it is true. It is not the fact that eddy currents are currently dissipating energy which creates the thermal noise, but that the mechanism is present and the noise source (the core) is at a non-zero temperature and therefore will have thermal fluctuations.
The coil integrates all the rate of change of flux passing through it. Heating a core does not provide a rate of change of flux through the core.

blue comments interspersed again..for clarification
Quote:
Originally Posted by DF96 View Post
Take a transformer (or a choke). Connect it to an amplifier or meter, so you can hear/measure noise. Just before you do this, measure the DC resistance and the AC resistance(measures all losses, wire and core eddies) at some appropriate frequency/ies. Measure the noise voltage, and heat the core while trying to avoid heating the windings (not easy, but it will take a while for heat to transfer between them so be quick). You should see the noise increase.(this is the crux of our disagreement, and is pretty much exactly the test I recommended with 4 different core types, a thousand posts or so ago..) Quickly repeat the DC and AC resistance measurements. (Here, you are driving eddy current losses on a hot core, and it's resistivity will now be different, this is a confounder that cannot be neglected..)Measure/estimate the core temperature. Measure/estimate the winding temperature (possibly from any rise in DC resistance). Calculate the thermal noise from the given resistances and temperatures (assume that winding resistance is unchanged from the DC value?). Compare with noise measurements.

Not an easy test(you are a master of understatement.. ), but it is the best I could think of to show the effect. The idea is to show that raising the core temperature increases noise, both by raising the corresponding resistance and raising the temperature of that resistance.
It would be better to do the 4 core test, as it changes in essence, the eddy loss component only. Honestly, what Scott mentioned is the better test...drive the coil using a variable sine, and look for sidebands.

Cheers, John
  Reply With Quote
Old 16th March 2011, 02:20 PM   #11404
diyAudio Member
 
john curl's Avatar
 
Join Date: Jul 2003
Location: berkeley ca
DF96: 'Eddy currents in Composite laminations' JAES Dec. '77, from IRE, Feb. 1936
  Reply With Quote
Old 16th March 2011, 02:22 PM   #11405
diyAudio Member
 
jneutron's Avatar
 
Join Date: Jan 2004
Location: away
Heres some good (if simplistic) papers from Dr. Sullivan et al, of Dartmouth.
I hope the links work.

note: Odd that he refers to an HP 4284 as a network analyzer, yet the manual 4 feet away from me says "precision LCR meter".

This paper has a gapped core 1500 turn #36 AWG measurement. Note that he discusses gapped cores as well in section IV. Gapped cores reduce the inductance, but more importantly, it increases the proximity losses of the wires that are close to the gap. He has some other papers on his site which detail optimum wire locations within a gapped core vs frequency, but that's something I'd recommend only for interested parties..


http://engineering.dartmouth.edu/ind...apers/sfdj.pdf



Here is a light paper describing the lowering of the relative permeability of round wire winding. This is basically a discussion of how the conductor reacts to the time varying flux the wire is embedded in. It is exactly what happens as well, when I put a half inch thick aluminum plate into the magnet gap..

http://engineering.dartmouth.edu/ind...apers/perm.pdf



Here is an easy read paper, the first page has some good pictures of the difference between skin effect and proximity effect..what I like is, it is extremely close to what occurs in my superconducting wires and magnets..very cool...

http://engineering.dartmouth.edu/ind...ers/litzcj.pdf

Cheers, John

ps...the use of the terms "simplistic" and "easy read" are most certainly tongue in cheek..hey, ya gotta have fun...

pps..note that MY referred papers are from this century...thank you very much....

Last edited by jneutron; 16th March 2011 at 02:25 PM.
  Reply With Quote
Old 16th March 2011, 02:38 PM   #11406
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Quote:
Originally Posted by jneutron
You are basically saying that there will be eddy losses despite there being no magnetic flux rate of change present.
There will be the mechanism. The flux will be changing due to thermal fluctuations in the core. The average flux (averaged over infinite time) will be zero. Average over any finite time and you get some net flux - the shorter the time the larger the net flux. You also need to average over all the lams, but the same principle applies. Each lam contributes a small amount of thermal power, so you add the power not the flux (flux adds as root mean squares).

I guess the ideal test would be a hot core and superconducting winding! That would guarantee no thermal noise from the winding itself, just the core.

I agree that the test with identical windings and different cores would be easier. The difficulty might be ensuring that lam thickness was the only variable.
  Reply With Quote
Old 16th March 2011, 02:46 PM   #11407
jam is offline jam  United States
diyAudio Member
 
jam's Avatar
 
Join Date: May 2001
Location: Auburn, CA, USA
John,

In my estimation (about life and cars) it is not so much about what you do..... but the important thing is that you look good doing it.

Your humble disciple.

Jam
  Reply With Quote
Old 16th March 2011, 02:47 PM   #11408
diyAudio Member
 
john curl's Avatar
 
Join Date: Jul 2003
Location: berkeley ca
You better get humble, grasshopper!
  Reply With Quote
Old 16th March 2011, 04:23 PM   #11409
diyAudio Member
 
jneutron's Avatar
 
Join Date: Jan 2004
Location: away
Quote:
Originally Posted by DF96 View Post
There will be the mechanism. The flux will be changing due to thermal fluctuations in the core. The average flux (averaged over infinite time) will be zero. Average over any finite time and you get some net flux - the shorter the time the larger the net flux. You also need to average over all the lams, but the same principle applies. Each lam contributes a small amount of thermal power, so you add the power not the flux (flux adds as root mean squares).
Average over any finite time?? Like picoseconds averaged over square nanometers?

Please detail exactly what net flux is available more than 100 nanometers from the surface of the core. Then detail exactly how that net flux is aligned with the core such that the coil sees a net flux passing through the windings.

Remember that the coil is not capable of generating an induced voltage as a result of infrared radiation. Yes, it can get hotter, but we're considering that as a considered/controlled entity. So you must provide a mechanism which produces time varying flux within the audio band that is positioned/focussed such that it travels through the coil to be picked up as a net rant of change of flux. Until you can provide that mechanism, the assumption that magnetic energy generated by thermal atomic motion within the core can create a magnetic field capable of pickup within the coil...is an unfounded assumption. (in my humble oinion..)
Quote:
Originally Posted by DF96 View Post
I guess the ideal test would be a hot core and superconducting winding! That would guarantee no thermal noise from the winding itself, just the core.
And the 4 kilowatt cryopumps...
Quote:
Originally Posted by DF96 View Post
I agree that the test with identical windings and different cores would be easier. The difficulty might be ensuring that lam thickness was the only variable.
When I first provided the mechanism of eddy current losses making noise, I explained in detail what the mechanism was, how it acted, and VERY importantly, I provided a test designed to expicitly prove or disprove the validity of my assertions. Scott provided the sideband content as a good metric, as I would have been left with looking at a high voltage sine for a hint of noise.

So far, I've seen no doable test which comes close with fewer confounders. Perhaps it would pay to do my test..

Cheers, John

Last edited by jneutron; 16th March 2011 at 04:52 PM.
  Reply With Quote
Old 16th March 2011, 04:54 PM   #11410
diyAudio Member
 
scott wurcer's Avatar
 
Join Date: Jan 2004
Location: cambridge ma
OK, I'll waste a little time on this today just to give us something tangible to talk around. Here is a spectrum of the open circuit noise of the worst tape head I could find (VHS audio track). DCR = 385 Ohms somewhere around -102dB on my plot. Ok where does the noise at 10k come from? It's almost 6-7K Ohms equivalent.

EDIT - Hate to dissapoint the "cores don't work through zero guys" but the noise has a perfect gaussian amplitude distribution even at this nano-volt level.
Attached Images
File Type: jpg tnoiz.JPG (77.1 KB, 170 views)
__________________
"The question of who is right and who is wrong has seemed to me always too small to be worth a moment's thought, while the question of what is right and what is wrong has seemed all-important."

Last edited by scott wurcer; 16th March 2011 at 05:05 PM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 09:20 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2