Need impressions of capacitors

[bzfcocon]

I'm sorry, maybe I've overreacted a bit.

On the positive side, I AM glad that somebody is making at least attempts to scientifically research on a matter that has been always subject to dispute. I am sure it was hard work, and, in spite of my criticism, it is the end result (=caps) that matter.

Thanks for making this info available - there's no way you can call that spamming !

 

[hermanv]

Many attempts have been made to quantify capacitor distortion. Here are a couple of links (scroll down about 2 pages) that go back decades. Nor are these the first.

http://stephan.win31.de/capdist.htm

In particular there is some excellent data in this article by John Curl and Walt Jung.

http://waltjung.org/PDFs/A_RealTime_Signal_Test_For_Capacitor_Quality.pdf

Note the dates on some of the references. I'm reasonably sure that Bell Labs looked into this matter in the 1920's when vacuum tube repeaters started to become common, especially as applied in the early trans-Atlantic telephone cables.

 

[bzfcocon]

I have to withdraw something, there is actually a picture that does show something useful: the frequency "doubling" effect.

However, while I have no reason not to trust the authors, I have to wonder about the size of the resonance peak: from the figure it's about 20dB, which I find very high. They do not say, however, for which kind of capacitor it was measured. Overall, I think they could have been a bit more specific on important issues.

 

[Curmudgeon]

Thanks for posting this. It is a bit skimpy on details for commercial reasons; as a result though, I will listen more to the boutique manufacturers who have paid a lot of attention to mechanical considerations.

 

[soongsc]

How do they connect the capacitor to the devices during the test ClarityCap made??

 

[roland bios]

Bump,

About the ClarityCap research. The full paper was presented at the AES convention in Amsterdam. The reactions were positive. The paper is under AES copyright, but i expect a more detailed public version soon.

Paul Dodds form ClarityCap explains the research and the outcome very well in this movie: (11 min)
Dodds on resonances

I will update asap if there's more info on this research.

Kind regards
Roland

 

[soongsc]

Bump,

About the ClarityCap research. The full paper was presented at the AES convention in Amsterdam. The reactions were positive. The paper is under AES copyright, but i expect a more detailed public version soon.

Paul Dodds form ClarityCap explains the research and the outcome very well in this movie: (11 min)
Dodds on resonances

I will update asap if there's more info on this research.

Kind regards
Roland

This is quite in line with the harmonics that I measured on various caps. It is very basic knowledge that plates with opposite charges will be attracted to each other.

 

[roland bios]

Hi,

But it's a new insight that this internal resonance is the mayor factor in sound differences*. Even more interesting is that these results were "proven" with ABX testing.

*according the research.

 

[soongsc]

Hi,

But it's a new insight that this internal resonance is the mayor factor in sound differences*. Even more interesting is that these results were "proven" with ABX testing.

*according the research.

There were two capacitors I sampled that are on Tony's site that were rateded like 7.5 or more somthing like that. We did attribute the difference in sound to winding tightness. It would be interesting to hear how the new series sound.

 

[Curmudgeon]

The Clarity work sounds like the most comprehensive. However, other manufacturers have come to similar conclusions; NorthCreek and Cardas come to mind.

 

[Salas]

Bump,

About the ClarityCap research. The full paper was presented at the AES convention in Amsterdam. The reactions were positive. The paper is under AES copyright, but i expect a more detailed public version soon.

Paul Dodds form ClarityCap explains the research and the outcome very well in this movie: (11 min)
Dodds on resonances

I will update asap if there's more info on this research.

Kind regards
Roland

Aha! Very nice video. At last some answers.

 

[roland bios]

Hi,

Finally some more info about the research from ClarityCap:
Full paper

This is as detailed as it's going to get...

Regards
Roland

 

[gedlee]

Appears to be a misleading report. No one could detect cap A from cap B in the initial blind ABX test, yet they had a preference in the later test. That is inconsistant, unless the final test used an electrolytic cap - they never do say what the comparison cap was. A bit too much smoke and mirors for my taste.

 

[critofur]

Re: Capacitors in crossovers

Originally posted by hermanv ...If you are one who has heard no difference between various interconnect cables, don't bother with capacitors, the differences are quite similar to cable variations [/B]

HAHAHAHAHAHAHAHAHAHAHAHA.....

Cables shouldn't make any sound, if they do, something is VERY wrong, and there's probably going to be a fire very soon.

Like, when we used the current probe instead of the voltage probe on our torroidal the other night... Doh!

 

[roland bios]

The way to get from zero to significant results is quite clearly descibed in the paper. Please point out the smoke and mirrors.

I find the new ABX method used expecially interesting. Let listeners use their own music, no time pressure. This may be of a more general interest far beyond this capacitor test.

 

[hermanv]

Re: Re: Capacitors in crossovers

HAHAHAHAHAHAHAHAHAHAHAHA.....

Cables shouldn't make any sound, if they do, something is VERY wrong, and there's probably going to be a fire very soon.

Like, when we used the current probe instead of the voltage probe on our torroidal the other night... Doh!

Capacitors also shouldn't have a sound signature, but they do, incidentally all cables are capacitors. This is the wrong forum for this discussion, there are many threads in this forum on this topic. If only it were as simple as you believe.

Cheers.

 

[Andre Visser]

Re: Re: Re: Capacitors in crossovers

If only it were as simple as you believe.
Cheers.

Nicely said.

 

[gedlee]

Please point out the smoke and mirrors.

I did. How can one have a preference between two caps which under blind ABX tests cannot be told apart?

And the second test is flawed because the listener always knows that A is A and B is B so that once they make up there mind - or guess - all future auditions are not blind. To be valid the caps A and B would have to be randomized from trial to trial with multiple trials performed in order to determine that differences could indeed be heard.

 

[soongsc]

Appears to be a misleading report. No one could detect cap A from cap B in the initial blind ABX test, yet they had a preference in the later test. That is inconsistant, unless the final test used an electrolytic cap - they never do say what the comparison cap was. A bit too much smoke and mirors for my taste.

Well, I once took some capacitors that were made by the same manufacturer, different lots, but the manufacturer claimed to be the same. However, they did not sound the same. I then went and graded them from listening, and did some harmonic distortion measurements using them in a low pass filter. Interestingly, there was quite well relationship between my grading and the measured results. The manufacturer then confessed the source of material was different. I think there is going to be some mytery as to how to make capacitors better.

I recall seeing an explanation from clarity cap indicating the foil internal vibration is one source causing the difference, but I have not yet determined a relation between tightness and sound quality. I suspect there are lots more things at play.

 

[roland bios]

And the second test is flawed because the listener always knows that A is A and B is B so that once they make up there mind - or guess - all future auditions are not blind. To be valid the caps A and B would have to be randomized from trial to trial with multiple trials performed in order to determine that differences could indeed be heard.

So how did they end up all liking "B" over "A"? It's not flawed it's another way of testing. And this other method is well described...I find it interesting they found results with this method which they could not find with ABX.