Testing audiophile claims and myths - Page 127

25ft of 16ga copper speaker wire with a 4 ohm load has less than .5dB of loss.

Line level applications have loads so high there is no appreciable resistive loss.

Level match due to cable loss really isn't an issue.

So what would be the loss of a headphone cables eg 8 foot of 24 gauge wire driving say a 50 Ohm headphone?
Edited by drez - 2/9/13 at 12:20pm

Less than 0.2dB.

 

I this I see your point.  So I guess any change is going to be less than 0.1 dB in which case inaudible or impossible to level match anyway?

Inaudible is the point.  

 

Your earlier comment re: cables not being level matched is for the most part unfounded, especially in line-level applications.  My point here is to show that a common length of 16ga speaker wire has an insertion loss too small to upset a level match situation, even at an impedance minimum of 4 ohms. The second point, the headphone application, is that that situation results in even less cable attenuation.  

 

Sorry, nothing personal, I just can't sit at let the concept of "cable loss" be thrown around without bringing it down to reality.

 

 

No problem I appreciate your input a lot, these are things I do not understand fully on a technical level. I don't mind being corrected if I learn something.

It would be conceivable that the impedance, capacitance, or construction of the cable (or insulator) could very minutely effect the frequency response of the cable, if I'm not mistaken, because impedance is frequency dependent.

Those differences in LRC values would have to be pretty far off to be anywhere near audible.

Inner Fidelity are doing some interesting tests of headphone cables, so far the differences are very slight, enough that issues with earpads seal could be distorting the results. I have to say I wasn't expecting any different, most sensible cables I have compared only possibly had minute differences, and yes my comparisons were uncontrolled so feel free to take that anecdote with a grain of salt. Still it was enough for me to have a preference one way or the other, which is certainly a very convoluted psychological construct if it were one.

The big key here is the driving source impedance. No common length of interconnect can have enough capacitance to affect frequency response unless the driving impedance is quite high. The days of high impedance outputs are long gone. Inductance of any amount that could affect frequency response is simply a physical impossibility without adding a deliberate and significant series inductor.

For speaker cables the surrounding source and load impedances completely swamp out anything the cable can provide in terms of L and C, with the possible exception of non-paired conductors with very large distances between them (like on opposite sides of the room) which creates a bit of L to contend with. The big factor in speaker wire is R, which is related to cross sectional size.

for completeness it should be pointed out that skin/proximity effect loss in heavy gage speaker cable with 4-8 Ohm speakers can be measurable by 20 kHz - but below audibility for most home wiring cases

 

for any weight wire that I'd want to attach to headphones the wire is much smaller than the skin depth at any audio frequency, and headphone Z is higher than loudspeakers - making the skin/proximity loss negligible

 

http://www.head-fi.org/t/517462/why-are-speaker-wires-frayed#post_7011124

 

 

[this thread is going in circles, I've posted the above link before]

Edited by jcx - 2/10/13 at 12:33am

 

Imo such answers should a) be better specified (0.000001dB is also less than 0.2dB) and b) be backed up by a formula, after all this is the 'Testing audiophile claims and myths' thread so if we all start shouting numbers, seemingly out of the blue, we're doing nothing but feeding the myth. Don't get me wrong, I suppose you did calculate this number somehow but it would be nice to see how so that others can repeat the calculation for their own setup if they'd want to.

 

It is not difficult to calculate, you can find the resistance of 1 foot of 24 gauge wire in a table like this, which happens to be 25.67 mΩ (note: this is the resistance of 1 foot of 24 gauge solid wire, or stranded wire that has equivalent cross sectional area of copper). So, the resistance of an 8 ft cable is ideally 0.02567 * 2 * 8 = 0.41072 Ω (it is multiplied by 2 because of the ground (return) wire). That causes 20 * log10(50 / (50 + 0.41072))  = -0.071 dB gain (attenuation) on a 50 Ω load.

Edited by stv014 - 2/10/13 at 3:11am

I suppose I did calculate it too, even if quickly and roughly, and thanks to stv014 for the example.   I almost did show the example, but considering the apparent audience here and their apparent lack of understanding, I could probably have published any formula and nobody who didn't already get it would know the difference anyway.  Didn't seem worth the trouble and time, I'll re-consider next time.  

 

But as to all shouting numbers...um...I don't hear numbers coming from the mythology camp, ever.  Because if they did make a claim like "standard cheapo wire produces an XdB roll off at Y-KHz" because of Z-effect, we'd just go prove them completely wrong simply and easily.  No, the number-shouting only comes from people with proof, even if the algebra isn't shown, it's easy to check. 

 

Everything you need to know is pretty much out there already, published, and possibly there's a web-based calculator that does most if not all of the job for you. 

 

So, you take me to task for not publishingthe example, yet no problem with somebody who says "Oooh, watch out for the capacitance, or impedance (we never touched on that one..) as if that were fact too. 

 

As much as my lack of "proof" may have upset you, people posting comment without Googling the concept first annoys me too.

 

Not completely. There is still some tube-based gear out there with output impedances in the thousands of ohms. But in the scheme of things, it's pretty rare.

 

se