Pjotr said:
Hmmm. According to my Webster's Ninth Collegiate:
"a retardation of the effect when the forces acting upon a body are changed (as if from viscosity or internal friction); esp : a lagging in the values of resulting magnetization in a magnetic material (as iron) due to a changing magnetizing force."
So it would seem that hysteresis has everything to do with time dependence. Which seems to perfectly describe dielectric absorption. Just as magnetic hysteresis is due to residual magnetization of magnetic materials, dielectric absorption is due to residual polarization of dielectric materials.
In other words, as you increase the magnetizing or polarizing force you create an initial magnetization or polarization of the material. When you bring the magnetizing force or polarizing force down to zero, the magnetization or polarization is something other than zero and requires a magetizing or polarizing force of some magnitude of the opposite polarity before the magnetization or polarization of the material becomes zero.
So, again, I don't see any difference between dielectric absorption and dielectric hysteresis.
se
What is going on here? Why so many glib opinions about cap properties? Learn about this stuff, folks! It just gets confusing to have people brush off some things as obvious, and talk about others in an incomplete way.
Now, before you all jump on me for good reason (good to you, at least ;-) ), let me put it into perspective.
First of all DA used to be VERY important with ANALOG computers. Remember them? They solved differential equations, very elegantly, with mathematical modeling of integration and differentiation by using caps in different ways. A small error in the cap threw off the results. Sometimes the output went bonkers!
Heck, they modeled the polystyrene cap as a potential error generator!
Now, lets look at other caps, you know: mylar, paper, oil, ceramic. They are pretty bad, 10-10000 times worse than polystyrene.
Now, did the manufacturers of these other caps advertise their problems? NO! We had to find them ourselves. We started publishing our results 25 years ago. You might say, why did 'we' care? That's Walt Jung, Dick Marsh, and me? Well, we designed phono stages that used caps to INTEGRATE the input phono signal. We had the same problem face us as the folks did with their analog computers, in the 50s and 60's.
Now, if you think about a wire, it is basically a cap in the transverse direction, and the electromagnetic signal that flows at nearly the speed of light, is mostly in the wire covering, which is a dielectric. How good are these coverings? Heard of any quality caps being made of that material? How about that doorbell wire that you got at the hardware store? Good enough for audio?
Now, most DA as measured in caps is LINEAR DISTORTION. I know, because I measured it hundreds of times. However, normal hysteresis, to my knowledge, is NONLINEAR. Are they really the same mechanism? I doubt it, and putting them in one classification can be confusing. Enough for now.
Now, before you all jump on me for good reason (good to you, at least ;-) ), let me put it into perspective.
First of all DA used to be VERY important with ANALOG computers. Remember them? They solved differential equations, very elegantly, with mathematical modeling of integration and differentiation by using caps in different ways. A small error in the cap threw off the results. Sometimes the output went bonkers!
Heck, they modeled the polystyrene cap as a potential error generator!
Now, lets look at other caps, you know: mylar, paper, oil, ceramic. They are pretty bad, 10-10000 times worse than polystyrene.
Now, did the manufacturers of these other caps advertise their problems? NO! We had to find them ourselves. We started publishing our results 25 years ago. You might say, why did 'we' care? That's Walt Jung, Dick Marsh, and me? Well, we designed phono stages that used caps to INTEGRATE the input phono signal. We had the same problem face us as the folks did with their analog computers, in the 50s and 60's.
Now, if you think about a wire, it is basically a cap in the transverse direction, and the electromagnetic signal that flows at nearly the speed of light, is mostly in the wire covering, which is a dielectric. How good are these coverings? Heard of any quality caps being made of that material? How about that doorbell wire that you got at the hardware store? Good enough for audio?
Now, most DA as measured in caps is LINEAR DISTORTION. I know, because I measured it hundreds of times. However, normal hysteresis, to my knowledge, is NONLINEAR. Are they really the same mechanism? I doubt it, and putting them in one classification can be confusing. Enough for now.
John, I don't follow your analogy. The proportion of the signal that flows through the cable capacitance is pretty low (cable capacitance is parasitic), whereas with coupling caps, it's just the opposite.
It's interesting that the guy who developed the tests you've used for caps, Scott Wurcer, doesn't think that wires are much of an issue, except in very special circumstances. Of course Wurcer thinks that everything else in high end audio is 99% BS, anyway, but he's a big freak about quality sound, a real music lover.
It's interesting that the guy who developed the tests you've used for caps, Scott Wurcer, doesn't think that wires are much of an issue, except in very special circumstances. Of course Wurcer thinks that everything else in high end audio is 99% BS, anyway, but he's a big freak about quality sound, a real music lover.
john curl said:
What's the difference between "normal hysteresis" and the hysteresis due to DA?
se
Say hello to Scott Wurcer for me. Haven't spoken to him for a few years. I use his op amps when I can. Use the AD797 in my ST THD analyzer front end. You know, the source of all my problems. ;-) I don't agree with Scott on this, but I know what I have to do to make good audio, Scott doesn't do it for a living like I do.
Steve Eddy said:
Normal hysteresis is i.e. what you can see at a comparator (electronic building block) with a small amount of positive feedback.
DA results in a frequency dependent phase shift. And who says this frequency/phase realationship is perfectly linear?
SY, we are talking here of very low levels of distortion, so why will the small parallel cap formed by the dielectric of the cable not be part of the signal path?
And oh eh if you google on dielectric+hysteresis a lot comes up but most related to piezo electric materials and thin film polymers.
Cheers
John: Scott is one of the smartest people I know. And a truly good guy, too. I hope to see him in the next month or so and will pass along greetings.
Pjotr, the capacitance is in shunt not series, so it has only a minor effect, simple HF rolloff, which doesn't happen until many, many times the highest audio frequency, or even more if the ciruit driving the cable has competently low source Z. The capacitive nonlinearities are down a couple orders of magnitude from that. So, if the effects exist at all, they're really miniscule, and frankly quite negligible compared to all the other nonlinearities floating around (like spekers, mikes, and rooms!). If you want to argue that .0001% harmonic distortion at 100kHz (say) is audible, well fine, but I haven't seen evidence of anything like that. Or frankly, any evidence of the audibility of any of the stuff proposed in this thread.
Pjotr, the capacitance is in shunt not series, so it has only a minor effect, simple HF rolloff, which doesn't happen until many, many times the highest audio frequency, or even more if the ciruit driving the cable has competently low source Z. The capacitive nonlinearities are down a couple orders of magnitude from that. So, if the effects exist at all, they're really miniscule, and frankly quite negligible compared to all the other nonlinearities floating around (like spekers, mikes, and rooms!). If you want to argue that .0001% harmonic distortion at 100kHz (say) is audible, well fine, but I haven't seen evidence of anything like that. Or frankly, any evidence of the audibility of any of the stuff proposed in this thread.
SY said:
Funny you mentioned that, because this thread is all about what can possibly be measured and not what is audible, contrary to most other threads 😀
Cheers 😉
Pjotr said:
Ok. So how does that hysteresis differ from the hysteresis due to DA?
Any reactive element or any other element which causes a time delay results in a frequency dependent phase shift.
DA results in hysteresis. It's the electric field equivalent of hysteresis in ferromagnetic materials which involves magnetic fields.
Never said it was perfectly linear. But that's not the point.
Ideally, DA is linear. And DA by its very nature results in hysteresis.
And you'll also find that the glossary at FaradNet uses DA, DH and soakage synonymously.
se
Pjotr said:
Actually this thread was about whether the distortion John has been measuring is being produced by the cables he's measuring or the equipment he's measuring them with. That is until all you threadjackers came along. 😀
se
Hi,
Which is their and your mistake for none of the terms are exact synonyms.
But be my guest, go on spouting incorrect information and I'll enjoy watching how you twist and turn to talk yourself out of it as usual.
It seems obvious to me that when you quote a reference on the net it's science and when others cite another reference it all become shady and not so scientific.
How convenient and utterly unscientific.
Cheers,😀
Which is their and your mistake for none of the terms are exact synonyms.
But be my guest, go on spouting incorrect information and I'll enjoy watching how you twist and turn to talk yourself out of it as usual.
It seems obvious to me that when you quote a reference on the net it's science and when others cite another reference it all become shady and not so scientific.
How convenient and utterly unscientific.
Cheers,😀
Hi,
I'd agree with that on first looks although there still is a nuance depending on what type of cap you have in mind.
Obviously soakage can have far more severe effects with electrolytics, when discussing the effect of DA with PTFE insulators it becomes a more subtle story.
In my ears I trust.😉
I'd agree with that on first looks although there still is a nuance depending on what type of cap you have in mind.
Obviously soakage can have far more severe effects with electrolytics, when discussing the effect of DA with PTFE insulators it becomes a more subtle story.
In my ears I trust.😉
Sy, you sure 'trust' the funniest guys! ;-) First, let me fill you in a bit.
When I first measured ceramic caps by the TEK method, 29 years ago, I found an unusual 'non-return to zero' state in cheap ceramic caps. NO OTHER cap had this obvious property. I take this to be hysteresis, because of what I saw on the scope screen. Other caps, like AL and tant can have equal or even more measured DA, which is a linear distortion and is modeled that way. DA, as measured by Scott Wurcer's differential subtraction test, seems to go away on continuous tone testing. You need an asymmetrical test signal to bring it out. Walt Jung and I wrote a paper on this 15 years ago.
Interestingly, Cyril Bateman is measuring 'non-linear' aspects of DA in some of his latest articles in 'Electronics World'. Maybe, he is on to something.
We are quibbling here.
When I first measured ceramic caps by the TEK method, 29 years ago, I found an unusual 'non-return to zero' state in cheap ceramic caps. NO OTHER cap had this obvious property. I take this to be hysteresis, because of what I saw on the scope screen. Other caps, like AL and tant can have equal or even more measured DA, which is a linear distortion and is modeled that way. DA, as measured by Scott Wurcer's differential subtraction test, seems to go away on continuous tone testing. You need an asymmetrical test signal to bring it out. Walt Jung and I wrote a paper on this 15 years ago.
Interestingly, Cyril Bateman is measuring 'non-linear' aspects of DA in some of his latest articles in 'Electronics World'. Maybe, he is on to something.
We are quibbling here.
Hi,
And that describes hysterisis perfectly. Something you'd experience in some caps.
I very much doubt to see this in your average I/C or filmcap but the same as DA it definitely is NOT.
SE's of this world take note.
Cheers,😉
And that describes hysterisis perfectly. Something you'd experience in some caps.
I very much doubt to see this in your average I/C or filmcap but the same as DA it definitely is NOT.
SE's of this world take note.
Cheers,😉
fdegrove said:
Show me that it's incorrect, Frank.
So far you haven't been able to cite a single source stating just what differentiates the terms nor have you offered any explanation yourself.
se
john curl said:
Perhaps because other caps don't use ferroelectric ceramics?
Since DA is a time domain phenomenon, why would you use continuous tone testing?
It's still there with symmetrical signals.
How does he know whether the nonlinearities have anything to do with DA?
se
- Status
- Not open for further replies.