cable dielectric cause of artificial sound


Hi folks, I would like to know what your opinion is about the following issue. About 90% of high-end cable manufacturers use PTFE as dielectric. Many of their cables sound much alike and they have a few of these characteristics in common: clean, relaxed and laid back sound but at the same time very dynamic (though a bit artificially), very quiet ("black background"), very good (also artificially) left/right separation. But I think albeit these traits, they tend to sound "technicolored", "sterile" and unengaging (lacking PRaT also). Some cable manufacturers are using bleached cotton as dielectric. These cables sound different: they have more natural dynamics, a mellower sound, more intimate soundstage, more tonal colors and so on. Are these differences mainly due to the dielectric material used? Why is for so many manufacturers PTFE still the ultimate dielectric for the use in audio cables?

Chris
dazzdax
!!!Transmission line theory does not apply at audio frequencies, unless your cable is a minimum of 500 feet long, and even then this is at 100 kHz and 1/10th a wavelength!!!

There is no debating this, not b/c I mention it, but b/c this is physical reality.
Some cable vendors would have you believe that characteristic impedance actually matters at audio bandwidth (do i hear any ee's in the Audience?). What total rubbish.
However, any given cable, coax, twin, twisted pair, will indeed have a characteristic impedance (this is purely a function of geometry and is nothing more than the ratio of voltage to current along the line).
This (Znot), however only becomes relevant when the cable length is longer than a fraction (usually 1/4 lambda) of the transmitted signal's frequency.
The signal frequencies passing through analog interconnects and speaker cables are effectively DC b/c the wavelengths are so long.
All this yak about characteristic impedance, load matching, and dielectric loss tangents (aka dissipation factor) applied to cable in the audio band is akin to applying special relativity when describing the point at which two cars will intersect on the highway (unless your car is a super hopped up red civic with a really fat tailpipe and has a mighty rocket engine), when good old distance = rate x time works perfectly every time....

I am not trying to disagree with anyone, and I did not invent TL theory, i'm just chiming in to remind, or point out, that all this talk about transmission line theory (dielectrics/insulation having appreciable influence beyond stopping our conductors from shorting out) just is not reasonable at analog audio frequencies.

I do think it is possible for such an ill designed cable to hit the streets such that it's basic RLC can affect the signal, but this is different than transmission line issues.

This type of cable might even have a coloration some find pleasant in their system.

for a real interesting (if you are participating in this thread I assume you are interested) and downright truthful (no bs) discussion of the matter check out:

http://www.audioholics.com/techtips/audioprinciples/interconnects/DielectricAbsorption.php
Dpac996: You and Audioholics are absolutely right!!! I recant all of my previous statements, experiences and chalk it up to my ignorance of the "facts". Thanks for clearing all of this up for me. Helping me to understand that all cables of reasonable construction and parts quality sound identical will make my life ( and maybe even yours ) SOOOOO much better. Sean
>
All this yak about characteristic impedance, load matching, and dielectric loss tangents (aka dissipation factor) applied to cable in the audio band is akin to applying special relativity when describing the point at which two cars will intersect on the highway
Tssk, tsk, Dpac996, really:).
What the audioholics article is/should be really implying is, for example, that connections are critical for, say RF (i.e. it simply won't work) and uncritical for audio frequencies (i.e. it WILL work -- but perhaps badly, i.e. with high losses and noise introduction).
Unfortunately, if you use a 50ohm cable on 75ohm connection you'll have reflections even at audio freq; if you create a highly capacitive interconnection, you may get oscillation, and will probably get attenaution of frequencies, even at audio frequencies; if you use additional conductors (say for shielding) you will change the electrical charactersitics of your connection and introduce shifts in the transmission; simple: if you use a thin conductor you introduce a higher resistance than with a thicker wire, and possible phase shift in lower frequencies vs higher frequencies -- even in the audible range (try it, it works!).

Etc, ad nauseam.

If I remember correctly, Sean works in RF applications, hence his comment:
Helping me to understand that all cables of reasonable construction and parts quality sound identical will make my life(...) SOOOOO much better
. How nice that would be!
Audioholics?
Since they don't believe that cables make a difference, they should just shut up about it and let us 'delusional' audiophiles enjoy our hobby.
Sean:

0"What the audioholics article is/should be really implying is, for example, that connections are critical for, say RF (i.e. it simply won't work) and uncritical for audio frequencies (i.e. it WILL work -- but perhaps badly, i.e. with high losses and noise introduction)."

THis is what you are saying. That article simply points out that dielectric loss tangents/DF are meaningless at audio band frequencies. It's not that an audio connection is uncritical, it's just unnecessary to view it like an RF transmission line. A ~50 kHz signal has a wavelength of 1000's of meters (lambda = propV / f ). Please someone, explain to me why anyone thinks this scenario should be treated the same as one where the conductor length itself is at or bigger than the wavelengths traveling on it.

That is why it's effectively DC (from a tl standpoint) on a wire of typical home cable lenghts.

"Helping me to understand that all cables of reasonable construction and parts quality sound identical"

1)I was trying to say that all cables do not sound the same b/c of , say reflections (vswr higher than 1), but rather lumped (rlc) chraracteristics. Inductnace and mutual inductance are based on geometry and small changes may manifest in different "sounding" (i think more like filtering) cables.

Gregm:
1"Unfortunately, if you use a 50ohm cable on 75ohm connection you'll have reflections even at audio freq"
Hmmm are you sure about this? please show me the article or data you have. Perhaps I am missing something

2"if you create a highly capacitive interconnection, you may get oscillation, and will probably get attenaution of frequencies, even at audio frequencies"

im with you on that. This has nothing to do with TL theory, this is basic lumped circuit analysis of the RLC.
Excessive capacitcane can make sources with high output Z unstable. No big deal. Once again the attenuation is due to 1st order RC filter. Not dielectric absorption or mistmatch

3"if you use additional conductors (say for shielding) you will change the electrical charactersitics of your connection and introduce shifts in the transmission;"
I think your partly right: you will add inductance. yes this changes the lumped characteristics. Any filter causes phase shifts but don't confuse this with V/I lag due to VSWR (voltage standing wave ratio).

4"simple: if you use a thin conductor you introduce a higher resistance than with a thicker wire, and possible phase shift in lower frequencies vs higher frequencies -- even in the audible range (try it, it works!)."
Once again you are describing the RLC componenets of a cable. I agree that changing these will change the sound.
Again this is not due to inherent mismatch b/t cable and source or sink.

We should not confuse source output impedance and sink input impedance with characteristic impedance as it applies in RF (freq >>100kHz) signal transfer.

And dudes: I am not trying to pick arguments with anyone. I am not the one who invented the theory. I am trying to relay the facts as i understand them. I sure might get confused myself in the description of them ( i try not but it's early and the java is empty) but they are out there for all to see. I am not a "measurement type" or one who sucks the life out of audio enjoyment. I just like to stick to the facts and separate myself from the rampant bullshit from marketing departments.