Good John Curl Article on Capacitors

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
actually there is a point or two to argue

dielectric adsorption is very well explained by a linear model, so speaking of nonlinear "dielectric adsorption" distortion is a little misleading

linear frequency response "distortion" - deviations from the pure C transfer function don't appear to be anywhere near audible thresholds - and I mean even quiet adapted ears noise threshold, much less the ~1% of signal level that double blind listening tests seem to indicate as the just noticeable frequency response difference

Bob Pease' Capacitor Soakage" article is a valuable information source:

http://www.national.com/rap/Application/0,1570,28,00.html

Dielectric Adsorption may correlate with other phenomena that are undesirable but audiophiles seem accept that it is proved that dielectric adsorption itself is a problem when the success of the linear model and calculable consequences of the linear model demonstrates that it is a audio non-issue

I'm not knocking Curl's comments in the interview so much as the continuing perpetuation in other audiophile write ups that the story is finished when they point to dielectric adsorption
 
I don't think it accurate to say dielectric absorption is "explained" by a linear model---it *is* modeled with a linear model.

Dielectric absorption is, IMO, quite audible, in fact is an obvious-as-the-day sonic differentiator between capacitors. Pease's article is quite old. A better article---series of articles, actually---and one that documents (sonically relevant) IMD and THD test results on different dielectric types is Cyril Bateman's 2002 Electronics World series on capacitor sound.
 
when the modeled response fits the measured data to a high degree of accuracy we generally say that the the model "explains" the phenomena - a colloquial usage - not necessarily a deep physics based causal explanation but a quite good enough "explanation" for engineers to move forward with

I seriously doubt anyone hears the physical phenomena of dielectric adsorption - you are claiming to perceive an audible difference in some circuit including a capacitor that is reproducing some musical or test signal in the audio frequency range - and then attributing the perceived difference in capacitor’s “sound” to the single variable of dielectric adsorption - this is not a logically watertight procedure when the capacitors undoubtedly differ in dielectric constant, loss, electrodes, construction, physical size, thermal history, water adsorption, lead material, wrapper color…

Since the linear circuit model of dielectric adsorption does “explain” the phenomena well enough for Bob to show a correction circuit for sample/hold applications that improves the lesser performing dielectric to Teflon performance levels then I think it isn’t to much of a leap to ask if the linear model “explains” reported audible differences – and it does not

This failure of a good model of the phenomena of dielectric adsorption to explain an observation of different audio perception of capacitors with differing dielectrics suggests to me that there may not a causal link to dielectric adsorption itself and that time would be better spent looking into other phenomena (that may just happen to correlate with dielectric adsorption)
 
Yes, yes, heard that before. If the clear-as-day audible factor is not dielectric absorption, it is, then, something in the neighbourhood of dielectric absorption, ie, something to which dielectric absorption very tightly correlates.

Bateman, a former capacitor engineer, would almost certainly disagree with your speculations regarding the sonic nature of dielectric-absorption-or-whatever-it-is-that-correlates-with-dielectric-absorption. Bateman, for his part, was, in writing his articles, just trying to be a good engineer furthering his science a bit.
 
single factor experiments are clearly out since we are talking different materials for the dielectrics

but a possible experiment would be to implement the linear multiple RC branch model of say a mylar capacitor with a literal implementation of the linear dielectric adsorption model using multiple polypropylene or Teflon Caps and quality resistors and abx compare them

the strong presumption from established double blind frequency response testing thresholds is that there would be no audibly perceptible differences for common linear coupling/filtering capacitor applications
 
john curl said:
The Bateman articles were very good. If someone can get the link(s) online, it would be appreciated.

Here's a link to a contrarian view ;)
http://www.electromagnetism.demon.co.uk/2644.htm

Maybe what our capacitors really need are little gold chips with quantumdots. I'm sure Machina Dynamica can design a "Cap Fixer" that magically expires after several capacitors, and magically knows which have and have not been "treated". Maybe we should just wind quantumdots in with the film.... I can see it now, Bybee or Machina Dynamica will start making "Quantum Caps".

There is entirely too much BS in high end audio.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.