Getting back to the original question.
With respect to coupling caps I don't think you can make a blanket statement that one type is superior to the other. The general assumption is that film and foil are superior; but I have certainly heard some pretty good metallised caps and some pretty dire foil ones.
I think the outcome is more due to materials selection, manufacturing processes and QA.
With respect to coupling caps I don't think you can make a blanket statement that one type is superior to the other. The general assumption is that film and foil are superior; but I have certainly heard some pretty good metallised caps and some pretty dire foil ones.
I think the outcome is more due to materials selection, manufacturing processes and QA.
Then, you're talking about non-linear distortion.
DA does not belong to that category.
Non-linear effects are caused by a change in permittivity as a function of the applied electric field, and they depend upon the absolute level of tension applied to the capacitor. DA remains the same at 1mV or 1KV.
If you use a 50V mylar capacitor at AC voltages of several volts, some non-linear effects will be noticeable, but if you use a 1000V part in the same conditions, the effects will practically vanish.
Walt jung's article Picking Capacitors part II has a description of how DA causes "distortion" of audio signals. Certainly to a layman like me it seems to be plausible. The attachment is just a small extract from the article. Both part one and part two are available on Walts site, where he gives quite a thorough (I thought) coverage of why you would want to use various types of capacitors in preference to others.
I've yet to read Cyril Batemans Understanding capacitors articles.
Tony.
I've yet to read Cyril Batemans Understanding capacitors articles.
Tony.
Attachments
Last edited:
DA doesn't give additional frequencies - its an entirely linear form of distortion. A capacitor with high DA can be modelled by adding long time-constant series RC networks in parallel with the original. Ceramics on the other hand can add distortion, but this is due to the dielectric having a voltage coefficient, not DA.
This is a purely subjective interpretation, and is totally unsupported by "hard" measurements.
Walt sort of implies that DA behaves like an audio delay line, which is not at all the case (in objectivist terms, again).
The actual representation of DA consists of distributed RC circuits.
That said, dielectrics having a high DA often have other types of imperfections, including non-linearity, which could explain a number of confusions.
Well yes Walt doesn't present any test that specifically demonstrates the effect of the DA on the transients in a music signal as he hypothesises (I imagine that with the measurement equipment available today this should be possible to prove or disprove). The only objective measurements are of the actual DA of various capacitor types using a test based on the mil standard test for DA.
Perhaps I should read the Cyril Bateman articles before further comment
edit: this article on Walt Jungs site does have some more "hard" tests, it is not stated however how it is determined that the effects measured are related to DA (which they state is the predominant factor) rather than other phenomenon...
Tony.
Perhaps I should read the Cyril Bateman articles before further comment
edit: this article on Walt Jungs site does have some more "hard" tests, it is not stated however how it is determined that the effects measured are related to DA (which they state is the predominant factor) rather than other phenomenon...
Tony.
Last edited:
Again, the problem is "what capacitor used where?" If cap A has 0R5 of ESR and cap B has 0R1 of ESR, there won't be any significant difference in a coupling use. Put the same caps in a high ripple current application (e.g., the input cap of a power supply filter) and the difference will be easy to measure- and possibly hear.
Trying to universalize things ("polypropylene is better than polystyrene" or "a lower DA cap will have lower distortion than a higher DA cap") is a sure path to confusion and audiophile legend. The requirements and important specs of a cap COMPLETELY depend on where and how it's used in a circuit. Repeat that three times.
Trying to universalize things ("polypropylene is better than polystyrene" or "a lower DA cap will have lower distortion than a higher DA cap") is a sure path to confusion and audiophile legend. The requirements and important specs of a cap COMPLETELY depend on where and how it's used in a circuit. Repeat that three times.
Yes, the real issue regarding distortion is what signal voltage is across the cap. In many cases this will be small, so distortion (even with a non-linear dielectric) will be small too. Coupling and decoupling caps, with appropriate values, might not be too critical. Problems can come from those capacitors which establish the LF and HF bandwidth limits, as they will carry signal voltage. The order of doing this can matter too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.