NO MODEL of a cap is an EXACT MODEL. Wake up, everybody! If you only have linear parts to model with, how are you going to model a nonlinearity?
Agreed, but things are more subtle than that: first, I do not think DA deserves a special treatment: it is just a loss mechanism, among many others; in fact, what is grouped under the appellation DA includes a number of more or less disparate effects.
What I have observed (OK, anecdotal, casual...) is that distortion in caps is always subordinate to a loss mechanism: capacitors having tan δ of 10e-4 or lower always have low distortion. This doesn't imply that relatively lossy caps have distortion: a 1.6KV mylar cap will have its regular 0.35% or so tan δ, but it will normally have a vanishingly low distortion when measured at 1Vrms (unless it has some flaw).
I have found PS caps having significant distortion, but they clearly had a problem: their loss factor was way too high for this dielectric (although still lower than a for a PET one).
This seems to mean that the permittivity must have an imaginary part for non-linear phenomenons to take place -but it is just a necessary condition, not a sufficient one-
Does anybody else have similar anecdotal (or even hard) evidence to share? Has someone found a capacitor that is at the same time almost loss less, yet heavily non-linear?
That said, even a perfect capacitor having vacuum as dielectric and infinitely stable electrodes will always have an intrinsic non-linearity, that stems from first principles, but I doubt it could be measured using standard DIY audio methods: it's more the stuff of CERN and similar institutions..
hope they are good models for distortion(s) sim..... you could say the same about the semiconductor data sheets (esp fet)..... they want to sell them so have the most accurate data (?). Maybe.... not so much.
-RNM
-RNM
They do react different than ideal models ...
Edit - in the case of new caps (solid) , the PC motherboard/mobile manufacturers need these
models. This seems to of driven the research.
OS
Edit - in the case of new caps (solid) , the PC motherboard/mobile manufacturers need these
models. This seems to of driven the research.
OS
Last edited:
???? Moving the goal post? I already talked about it recently........ I started with this inquiry, as I already stated, from a Berkeley researcher who did the cap comparison tests. before I did my test. [he thought it was DA.] So I made the test my way. etc Pls pay closer attention.
What was in the past is still in the past apparently here...... the non-linearities lead to servo and cap elimination. So, yes no one does it now.
Or do they.... I just said yesterday it is still being done A LOT... the Marantz PA 7025 (CFA) is newish and still uses polar electrolytic caps for coupling..... still a popular old way to do things.
Just not in the high-end any more because? ........ WHy is that? Cost isnt an issue at the high-end.
THx-RNMarsh
Last edited:
Factoid --- to get low or near zero Temp Co ..... 2 different dielectrics are used.... plus and neg TC. Nearly zero TC can be made that way.
-RNM
-RNM
Cute. Nice try. No cigar. Only conclusion is from smoking gun from tests results.
The DA comes closest to listener's description of the non-linearities they where compared to no cap (polar electro only)
-RNM
Last edited:
Guys, we researched this stuff half to death, 30-40 years ago.
It is like you guys just woke up from a 30 year old nap and are asking the same questions,(and giving the same opinions) of what we researched 30+ years ago. Get up to speed, guys, and don't think that your college education has taught you everything about engineering. '-)
It is like you guys just woke up from a 30 year old nap and are asking the same questions,(and giving the same opinions) of what we researched 30+ years ago. Get up to speed, guys, and don't think that your college education has taught you everything about engineering. '-)
Ooh just like clock makers discovered in 1726 and used ever since in various fields.
Cute. no cigar for you either. If someone had long a go a dielectric nonlinearity model which matched any and all tests/measurements, we could be done with this.
Maybe we just decide by voting on it --- the most this reason or that reason...... winner take all.
-RNM
Maybe we just decide by voting on it --- the most this reason or that reason...... winner take all.
-RNM
Last edited:
If you're sincerely curious, there's a rich literature on the dielectric properties of solids. A few minutes of searching will get you the names of standard texts which you can study.
Well, SY, why didn't we hear from you 30+ years ago? What you are trying to do is to discredit hi end audio, and any improvements that we have made to audio amps in general, over the last 40 years. Why you want to do this is beyond me, except that it might make you more secure in your own personal knowledge base (apparently derived from college).
"Silence between notes" is not silence, but the sound of the environment in which the recording was made, the residual of all the echos from the just done music making, and the intrinsic "noise" of the recording space itself. This is what separates the men systems from the boys, and is made clearer by reducing the distortion of the low level information - IME conditioning of the system, that is, longer term, vigorous running of a system always improves this, and having the capacitors reach an equilibrium of operation with respect to parasitic behaviours seems to be part of this ...
- Status
- Not open for further replies.