Capacitor rating vs VDC rating

[carlspeak]

Recapping is certainly a part of what, I presume, some folks who visit here and shop PE do.
I've recently completed a 6 pg. .pdf document on the subject above. Unfortunately, it's too large to attach here. However, you can visit here
http://www.classicspeakerpages.net/I...showtopic=5828
and download it.
What inspired my effort was an interest in Tony Gee's cap ratings posted on his website.
Also in my document are many links to other cap evaluations besides Tony's and other subjectivists vs objectivists blogs on the www.
As you may be aware, numerous discussions have gone on regarding whether or not boutique caps are worth the price and effort. This document looks at the problem from my view - that of an industry outsider who simply saw something in Tony's ratings and others which may or may not have relevance.
Read it and decide for yourself.

[dcibel]

I'll stick with Bateman's document as a reference:
http://www.scribd.com/doc/2610442/Capacitor-Sound

Bateman wrote a textbook or two that I used in school.

I understand that Tony is well respected in audio land, but the method's used for determining that something is better or worse when compared so something else are not very good in my opinion. At least he explains on his page that he just puts whatever cap in whatever speaker he has around at the time and makes that evaluation (that is, if anyone reads the top paragraph on his cap test page), so at least he's honest about the method.

The problem is that the method is not scientific at all. Too many variables change, testing is not blind, double-blind, whatever. The end result is that I can only take these type of ratings with a grain of salt. Something like Bateman's article which has accompanying theory and measurements, that is of great use.

In addition, I'm fairly certain most DIYers have come to the conclusion that money spent on a fancy expensive cap is a waste. Spending the same money on a better speaker, and executing a good design makes a real difference that anyone can hear. If one has money to burn however, then why not buy some Janzen superior caps...

[Deward Hastings]

Missing from many discussions of "capacitor sound" is a discussion of plausible mechanism, and without mechanism it is difficult to distinguish between 'real" and "imagined" differences, because no objective test is indicated. Bateman often obliquely and sometimes directly suggests that there is a testable mechanism . . . and although he doesn't identify it in exactly these words it is capacitance modulation by polarizing voltage.

The effect is particularly obvious in electrolytic capacitors, but can be found in some others. It is most particularly obvious in blocking capacitors when the polarity across the capacitor changes (what Bateman calls the "diode effect"). This effect is rarely seen in vacume tube circuits, where coupling capacitors typically have high potential on one side and are near ground potential on the other, but in transistor circuits (and crossovers) coupling capacitors typically see only a small DC offset and a large signal can (generally does) reverse polarity across the capacitor. Small signals are also typically found close to zero polarization, where surface and boundry layer effects are most prevalent.

In the absence of DC bias a high level low frequency signal constantly varies capacitor polarization, and thus capacitance, and produces the measurable and sometimes audible intermodulation distortion that Bateman shows. The classic SMPTE IM test may be as good as any for measuring this effect. Worst case would be (relatively) low level high frequency sounds riding over loud bass . . . just what we commonly have in music.

With regard to different voltage ratings (in film caps) sounding different it is at least possible that the thicker dielectric used to obtain the higher voltage rating diminishes the (relative) importance of surface polarization effects in favor of the non-variant bulk properties of the dielectric film.

[dcibel]

Somewhere on Joe Rasmussen's Elsinore discussion thread there is some talk about a crossover modification to apply DC bias to the capacitors, to essentially remove the effect you are describing. It looks like it would do the trick on paper, but would require something like a 9V battery to be hooked up to the speaker, or external power supply. I can see why this change would not be easily adopted commercially.

[Face]

Vandersteen used a battery biased crossover at one time.

[Æ]

I tried to download the article, sure wish I could. It looks to be a very informative reading. I tried to print/save it, but it wouldn't work.
Would you happen to know where I could find a .pdf of it?
Thanks

I'll stick with Bateman's document as a reference:
http://www.scribd.com/doc/2610442/Capacitor-Sound

Bateman wrote a textbook or two that I used in school.

I understand that Tony is well respected in audio land, but the method's used for determining that something is better or worse when compared so something else are not very good in my opinion. At least he explains on his page that he just puts whatever cap in whatever speaker he has around at the time and makes that evaluation (that is, if anyone reads the top paragraph on his cap test page), so at least he's honest about the method.

The problem is that the method is not scientific at all. Too many variables change, testing is not blind, double-blind, whatever. The end result is that I can only take these type of ratings with a grain of salt. Something like Bateman's article which has accompanying theory and measurements, that is of great use.

In addition, I'm fairly certain most DIYers have come to the conclusion that money spent on a fancy expensive cap is a waste. Spending the same money on a better speaker, and executing a good design makes a real difference that anyone can hear. If one has money to burn however, then why not buy some Janzen superior caps...

[dcibel]

No, sorry. I believe before it was recently posted at that link, it was something you had to pay for. I grabbed the link from a post at HTGuide.