Re: DA AND DH.
Dielectric hysteresis is reduced as the loss angle decreases and becomes zero when the loss angle is zero.
Since dielectric hysteresis is reduced as the loss angle decreases, how does that make it sufficiently clear to see that DA and DH are not one and the same?
se
fdegrove said:
Dielectric hysteresis is reduced as the loss angle decreases and becomes zero when the loss angle is zero.
Since dielectric hysteresis is reduced as the loss angle decreases, how does that make it sufficiently clear to see that DA and DH are not one and the same?
se
Any hot shots out there? Can any of you do a circuit simulation?
If so, then try this: Make a differential subtractor with an ideal instrumentain op amp simulation. Cap couple both inputs, and drive them from the same signal. Have equal value R's to ground on each input. As some of you probably realize, if the caps are the same value and create the same RC time constant on both the = and - input, then they will cancel out.
Now create a model of a moderate DA capacitor by adding series RC's in parallel to one of the input caps. You can find this data by looking around on Pease's soakage article. Then try different test signals at the input, after you have slightly adjusted one side for the best null with the test signal that you are using. Try a sine wave, try a sine wave sweep, try a square wave.
Then try an asymmetrical pulse. Note the difference. Report it here. Test results should be in by the end of the week. Hint: you should find a big difference between symmetrical input signals and asymmetrical input signals.
If so, then try this: Make a differential subtractor with an ideal instrumentain op amp simulation. Cap couple both inputs, and drive them from the same signal. Have equal value R's to ground on each input. As some of you probably realize, if the caps are the same value and create the same RC time constant on both the = and - input, then they will cancel out.
Now create a model of a moderate DA capacitor by adding series RC's in parallel to one of the input caps. You can find this data by looking around on Pease's soakage article. Then try different test signals at the input, after you have slightly adjusted one side for the best null with the test signal that you are using. Try a sine wave, try a sine wave sweep, try a square wave.
Then try an asymmetrical pulse. Note the difference. Report it here. Test results should be in by the end of the week. Hint: you should find a big difference between symmetrical input signals and asymmetrical input signals.
DH, possibly some references
(Shortcut: possibly relevant references at the end of post)
In a forum discussion with a lot of people participating, there
is always confusion about who says what and who thinks what
and we tend to lump people together into two or, at most,
three different camps to make things manageable. This is
of course always unfair, since often there are as many opinions
and standpoints as there are individuals.
Just to clear it out, I have not made any claim about whether
there is or is not such a phenomenon as dielectric hysteresis
and which is distinct from dielectric absorption. I have just
commented that it there Í have found no support for such
a phenomenon in those references and claims made about
it so far. I know of course, I could have made more extensive
web searches than just searching for the exact phrase, and
some grammatical variations of it, that Franks suggested, but
am I to find proof of his claims??? The main reason I posted
my fails results was that I found them quite funny in this
context.
Since I anyway can't do much useful in the mornings until I
have had more coffee than is good for me I decided to do
some more web search on the topic. I didn't wade through all
of what came up but let me give you the three possibly relevant
pointers pointers I found:
1) The quote "the dielectric hysteresis decreases as the loss angle of the dielectric increases" which Frank failed to give a
reference to can be found in this patent application
http://www.nia.org/timeline/text/1708038.htm
I didn't read much of it, since I hate reading patent applications
and they are anyway often just unsubstantiated claims.
2) I found a database entry for a IEEE symposium paper, but
no link to the paper, with the title "Apparent dielectric hysteresis without significant dipolar movement?" which might be of
interest, since DA requires dipolar movement. I don't have the
time or patience to try digging up the paper right now, at least.
http://fodok.jku.at/fodok/publikation.xsql?PUB_ID=9271
3) A thesis abstract having "dielectric hysteresis" among the
keywords. There is no indication in the abstract, however,
what the term means or if it is different from DA. Just like paper
2, however, it is from Austria, so maybe they are related.
http://www.arcs.ac.at/dissdb/rn038176
Now I've done my share of the work for the moment, or more
than that, so now I suggest somebody else go dig up these
papers if they care. I would personally start with #2, which
seems most likely to shed some light on the issue.
Edit: Considering recents posts by Frank and John, I should add
that my assumption has all the time been that DH is supposed
to appear not only in ceramic and electrolytic caps, which we
already know behave in peculiar ways of their own.
(Shortcut: possibly relevant references at the end of post)
In a forum discussion with a lot of people participating, there
is always confusion about who says what and who thinks what
and we tend to lump people together into two or, at most,
three different camps to make things manageable. This is
of course always unfair, since often there are as many opinions
and standpoints as there are individuals.
Just to clear it out, I have not made any claim about whether
there is or is not such a phenomenon as dielectric hysteresis
and which is distinct from dielectric absorption. I have just
commented that it there Í have found no support for such
a phenomenon in those references and claims made about
it so far. I know of course, I could have made more extensive
web searches than just searching for the exact phrase, and
some grammatical variations of it, that Franks suggested, but
am I to find proof of his claims??? The main reason I posted
my fails results was that I found them quite funny in this
context.
Since I anyway can't do much useful in the mornings until I
have had more coffee than is good for me I decided to do
some more web search on the topic. I didn't wade through all
of what came up but let me give you the three possibly relevant
pointers pointers I found:
1) The quote "the dielectric hysteresis decreases as the loss angle of the dielectric increases" which Frank failed to give a
reference to can be found in this patent application
http://www.nia.org/timeline/text/1708038.htm
I didn't read much of it, since I hate reading patent applications
and they are anyway often just unsubstantiated claims.
2) I found a database entry for a IEEE symposium paper, but
no link to the paper, with the title "Apparent dielectric hysteresis without significant dipolar movement?" which might be of
interest, since DA requires dipolar movement. I don't have the
time or patience to try digging up the paper right now, at least.
http://fodok.jku.at/fodok/publikation.xsql?PUB_ID=9271
3) A thesis abstract having "dielectric hysteresis" among the
keywords. There is no indication in the abstract, however,
what the term means or if it is different from DA. Just like paper
2, however, it is from Austria, so maybe they are related.
http://www.arcs.ac.at/dissdb/rn038176
Now I've done my share of the work for the moment, or more
than that, so now I suggest somebody else go dig up these
papers if they care. I would personally start with #2, which
seems most likely to shed some light on the issue.
Edit: Considering recents posts by Frank and John, I should add
that my assumption has all the time been that DH is supposed
to appear not only in ceramic and electrolytic caps, which we
already know behave in peculiar ways of their own.
Hi,
Well done, Christer.
I suppose it's worthwhile to actually do more than just a quick glance at the first page of the Google results.
Which is what I did and it's rather time consuming as you may now realise yourself.
Even more so when you actually read the interesting references.
So, may I suggest a few more coffees and some reading of the references you've just quoted?
Cheers,😉
Well done, Christer.
I suppose it's worthwhile to actually do more than just a quick glance at the first page of the Google results.
Which is what I did and it's rather time consuming as you may now realise yourself.
Even more so when you actually read the interesting references.
So, may I suggest a few more coffees and some reading of the references you've just quoted?
Cheers,😉
Frank,
You don't have to tell me it is time consuming to seach the web,
I have known that for years. 🙂
The thing is, you made the claim, so you have supposedly found
papers supporting your claims, so why not point us to them?
The burden of proof is usually with the one who makes the
claim.
Is any of the three references I found worth reading, for instance?
Don't have time for more coffee or reading for the moment
(no I don't mean dipole moment as in DA 🙂 ).
You don't have to tell me it is time consuming to seach the web,
I have known that for years. 🙂
The thing is, you made the claim, so you have supposedly found
papers supporting your claims, so why not point us to them?
The burden of proof is usually with the one who makes the
claim.
Is any of the three references I found worth reading, for instance?
Don't have time for more coffee or reading for the moment
(no I don't mean dipole moment as in DA 🙂 ).
Hi,
Would I dare to make the claim if I didn't know what I was talking about?
Pointing to references is no guarantee that people will actually read them, making them look them up themselves will at least teach them where and how to find them.
Sounds arrogant, I know, but sometimes it's the best way to go.
The Wilson patent is really old but worth reading.
The Austrian ones I vaguely remember as being both more or less the same but I'll need to reload them and see what's in them.
This one I found worth going through as it's more related to audio issues:
WEBER
Cheers,😉
Would I dare to make the claim if I didn't know what I was talking about?
Pointing to references is no guarantee that people will actually read them, making them look them up themselves will at least teach them where and how to find them.
Sounds arrogant, I know, but sometimes it's the best way to go.
The Wilson patent is really old but worth reading.
The Austrian ones I vaguely remember as being both more or less the same but I'll need to reload them and see what's in them.
This one I found worth going through as it's more related to audio issues:
WEBER
Cheers,😉
Hi,
Reference 2) is not worthwhile, 3) is a summary of a dissertation and is of interest to me as it treats another one of my pet topics related to audio wire namely dielectric barrier microdischarges.
Which is probably the origin of the "micro diode" theory blurb.
Cheers,😉
Reference 2) is not worthwhile, 3) is a summary of a dissertation and is of interest to me as it treats another one of my pet topics related to audio wire namely dielectric barrier microdischarges.
Which is probably the origin of the "micro diode" theory blurb.
Cheers,😉
The dipolar motion made me wonder about temperature dependence of DA- how much does it reduce going from (say) 60 degrees C down to 20 degrees C? Can one improve cap performance by intensive cooling? I would think that the T-dependence would follow a Boltzman law below the Tg of the dielectric, so this could be a significant effect ("significant" in the context of the miniscule effects we're discussing).
John, did you do any kind of gauge repeatability and reproduceability study on your test setup for the wire measurements? Can you share the results?
John, did you do any kind of gauge repeatability and reproduceability study on your test setup for the wire measurements? Can you share the results?
Hi,
Possibly so but wouldn't that also change the capacitors' value?
I'm thinking about the TKc factor here.
Cheers,😉
Possibly so but wouldn't that also change the capacitors' value?
I'm thinking about the TKc factor here.
Cheers,😉
Yes, though at a much slower rate than the change in DA, since the dielectric constant doesn't involve molecular or domaine motion, just polarizability.
john curl said:
Can I propose an alternative experiment, as follows?
1. Feed an input signal (of whatever kind), and note the output signal.
2. Analyse the frequency spectrum of the input signal - note which harmonics are present.
3. Similarly, analyse the frequency spectrum of the output signal.
4. Note and report whether any frequencies are present in the output signal which are not in the input signal.
The point of this experiment is that you don't need a simulator. Maths will simply show that the only frequencies present in the output were present in the input to start with, no matter how many parasitic capacitors, resistors and inductors you add.
Why am I bothered? Well, if DA is completely and accurately modelled by parasitic R's and C's, it cannot be the cause of any frequencies present in the output which weren't in the input (which is what I understand by terms such as nonlinearity and distortion). Its effects can be completely defined with a few simple, easily measurable parameters (e.g. frequency response). In other words, weird measurement techniques won't tell you anything that straightforward ones don't.
If DA can't be modelled like this (and, for instance, magnetic hysteresis effects in iron-cored inductors can't be modelled with parasitic R's and L's) then it's more interesting. If anyone has measurement results which can't be made to fit an R/C network model, let us all know - this is definitely new territory.
Cheers
IH
Re: DH, possibly some references
Yes, and what I responded with, which was the exact opposite of what Frank had stated:
Dielectric hysteresis is reduced as the loss angle decreases and becomes zero when the loss angle is zero.
Is from the same document.
By the way, you should read it. Even though it's nearly 80 years old, much of what's said in it wouldn't be out of place in some "white paper" written today by a high-end cable manufacturer.
se
Christer said:
Yes, and what I responded with, which was the exact opposite of what Frank had stated:
Dielectric hysteresis is reduced as the loss angle decreases and becomes zero when the loss angle is zero.
Is from the same document.
By the way, you should read it. Even though it's nearly 80 years old, much of what's said in it wouldn't be out of place in some "white paper" written today by a high-end cable manufacturer.
se
fdegrove said:
Did you even read the one you copied this statement from:
The dielectric hysteresis decreases as the loss angle of the dielectric increases.
If you had, you'd have found that it also contains this statement:
Dielectric hysteresis is reduced as the loss angle decreases and becomes zero when the loss angle is zero.
How do you explain these two diametrically opposed statements?
se
You know, after all this sqirmishing around, one can only conclude that neither John C, Frank or Chirster have any clue what they are talking about on these subjects.
I think this thread has degenerated to the the level of being disgusting.
It would be a major step forward if you all would start to act like grown-ups. But then again, I've been disappointed before.
Continue your mental, content-less masturbation, but without me. Which isn't a loss anyway.
Jan Didden
I think this thread has degenerated to the the level of being disgusting.
It would be a major step forward if you all would start to act like grown-ups. But then again, I've been disappointed before.
Continue your mental, content-less masturbation, but without me. Which isn't a loss anyway.
Jan Didden
Hi all,
I do not know DA but I study hysteresis phenomena from magnetic ones to those associated to
the adsorption of water by a porous media. I think, IMHO, that if this hysteresis is a linear one
the name hysteresis is incorrect.
Hysteresis is a nonlinear phenomenon, in some sense it is two times non-linear.
when I have a linear system ( e.g. a circuit containing only resistance, caps., inductors
and linear active devices ) I can describe it with a system of linear differential equations
with constant coefficients. When some devices are non linear the equation in the system become nonlinear, that is I can read terms, in the equations, that are a nonlinear function of the state variables (e.g. voltages and currents ). From here the terms “nonlinear system”, “non linear distortion”, etcetera.
When these nonlinear terms show discontinuities or are governed by “if … then” rules then
hysteresis may arise, or not. It may depend or not on freq. but certainly the behavior of a hysteretic system depends on its history.
Federico
I do not know DA but I study hysteresis phenomena from magnetic ones to those associated to
the adsorption of water by a porous media. I think, IMHO, that if this hysteresis is a linear one
the name hysteresis is incorrect.
Hysteresis is a nonlinear phenomenon, in some sense it is two times non-linear.
when I have a linear system ( e.g. a circuit containing only resistance, caps., inductors
and linear active devices ) I can describe it with a system of linear differential equations
with constant coefficients. When some devices are non linear the equation in the system become nonlinear, that is I can read terms, in the equations, that are a nonlinear function of the state variables (e.g. voltages and currents ). From here the terms “nonlinear system”, “non linear distortion”, etcetera.
When these nonlinear terms show discontinuities or are governed by “if … then” rules then
hysteresis may arise, or not. It may depend or not on freq. but certainly the behavior of a hysteretic system depends on its history.
Federico
Wake up folks, DA is usually defined and modeled as a LINEAR DISTORTION compared to nonlinear distortion, which is measured by harmonic or IM distortion measuring instruments. DA can still be many percent of the audio signal per cap. Sine wave analysis will hide DA, so a cap with, let's say 5% DA, will have less than .0005% harmonic distortion. What is that ratio? Quiz on Friday.
IanHarvey said:
Quite so. Even if your waveform is asymmetrical.
Correct.
Though alterations in frequency and phase response is considered "linear" distortion.
se
- Status
- Not open for further replies.