Imaginary scenario and I don't have big technological skills!
I need to get speaker cables and decide to use the AWG table to choose my copper wire, I want the 100KHZ of my SACD to reach the 4 ohms speakers.
At the table starting with the column Max frequency for 100%... 107khz-AWG26-max amps 0.361 so aprox. half a Watt.
I need to get speaker cables and decide to use the AWG table to choose my copper wire, I want the 100KHZ of my SACD to reach the 4 ohms speakers.
At the table starting with the column Max frequency for 100%... 107khz-AWG26-max amps 0.361 so aprox. half a Watt.
I did the cables again de 20 + and the 20 - all side by side and with 2.2 meters my meter reads 270 pf
Some "ethernet" cables are designed for HD media transfer where the skew between pairs can effect the picture, the same twist rate helps this.
Other wise im in the whats wrong with zip cord camp, the 100khz is more audiophoolery, most mics dont go over 25khz.
Other wise im in the whats wrong with zip cord camp, the 100khz is more audiophoolery, most mics dont go over 25khz.
Sorry but...
1) You can't hear frequencies that high.
2) Your speakers cannot reproduce frequencies that high.
3) Your source material almost certainly doesn't contain frequencies that high.
4) Most audio electronics are not designed for frequencies that high, and some could be driven into excessive distortion.
4) As pointed out by others it is possible that speaker cables with higher reactive components can cause amplifier instability and elevated distortion and noise.
5) Conventional, well made speaker wire of adequate current carrying capacity works just fine without adding extra problems.
Mike
Totally agree. 100khz is not something to worry about.
Me, I would only worry about settling time variations due to the line to load ratio vs frequency.
That is why I would only recommend dropping the speaker cable impedance to roughly 25 ohms.. Its a reasonable middle range of speaker impedance variations.
That z is basically 4 twisted pairs. However, I would never go below #14 awg for the four pairs.
Me personally, imaging is of no concern as I use music as background or high power PA, neither of which is real finess nor sweet spot listening.
But if I were worried, I'd use 20-25 ohm cables or put the amp in the cabinet.
Jn
Ps.. some recording mics do 40 kHz. No experience with that content.
Me, I would only worry about settling time variations due to the line to load ratio vs frequency.
That is why I would only recommend dropping the speaker cable impedance to roughly 25 ohms.. Its a reasonable middle range of speaker impedance variations.
That z is basically 4 twisted pairs. However, I would never go below #14 awg for the four pairs.
Me personally, imaging is of no concern as I use music as background or high power PA, neither of which is real finess nor sweet spot listening.
But if I were worried, I'd use 20-25 ohm cables or put the amp in the cabinet.
Jn
Ps.. some recording mics do 40 kHz. No experience with that content.
Lamp cord is among the best speaker cables.
The only thing that matters is low wire resistance. All other characterics are mostly BS.
The only thing that matters is low wire resistance. All other characterics are mostly BS.
Ok I was just trying to give an example why I think 1 pair of regular copper multistrand speaker cable is not the best option.
Just to be clear I am almost deaf above 10khz, but the standard for SACD is 100khz and I'm just consulting the AWG table.
I'm not an expert and looking at the standards for ethernet cables and comparing with AWG table (both for a lot longer cables that the ones we are using) I get questions I'm not able to answer, 2 examples:
- There are ethernet POE type 4 certified cables (960mA in each pair) made with AWG 28 and AWG table says max 226mA (aprox 4x).
- The AWG table says that AWG 23 as a max frequency for 100% skin depth of 53KHz if we look at cat7 standard or the picture of the one I'm using they are certified for 500-650MHz (aprox 10 000x that frequency).
Reading this thread I have another question, if we look at the differences between typical construction of shielded cat 6 and cat 7 in cat 6 we have a plastic "cross" core that is intended to maintain a minimum distance between the cables to reduce crosstalk and in the cat 7 they replaced it with foil that does almost nothing and as a side effect getting the pairs closer and yet they raised the max frequency from 250MHx to 500MHz how?
Just to be clear I am almost deaf above 10khz, but the standard for SACD is 100khz and I'm just consulting the AWG table.
I'm not an expert and looking at the standards for ethernet cables and comparing with AWG table (both for a lot longer cables that the ones we are using) I get questions I'm not able to answer, 2 examples:
- There are ethernet POE type 4 certified cables (960mA in each pair) made with AWG 28 and AWG table says max 226mA (aprox 4x).
- The AWG table says that AWG 23 as a max frequency for 100% skin depth of 53KHz if we look at cat7 standard or the picture of the one I'm using they are certified for 500-650MHz (aprox 10 000x that frequency).
Reading this thread I have another question, if we look at the differences between typical construction of shielded cat 6 and cat 7 in cat 6 we have a plastic "cross" core that is intended to maintain a minimum distance between the cables to reduce crosstalk and in the cat 7 they replaced it with foil that does almost nothing and as a side effect getting the pairs closer and yet they raised the max frequency from 250MHx to 500MHz how?
Don't worry about skin depth. For your application, it doesn't alter the performance. A #10 conductor will skin at lower frequencies, but the amount of copper still being used in conduction will be very large.
The only worry about skin depth is if you are running very high currents at very high frequencies. Audio does not reach that regime.
For a speaker wire, the major concern is end to end resistance of the conductors. Don't worry about 100Khz performance of the wire, a 4 ohm speaker will not be capable of drawing any current at 100 Khz even if humans could here it.
If you are pushing any power into a 4 ohm load, you should just go with #14 or #12. Even #16 lamp cord is sufficient at lower power levels.
jn
The only worry about skin depth is if you are running very high currents at very high frequencies. Audio does not reach that regime.
For a speaker wire, the major concern is end to end resistance of the conductors. Don't worry about 100Khz performance of the wire, a 4 ohm speaker will not be capable of drawing any current at 100 Khz even if humans could here it.
If you are pushing any power into a 4 ohm load, you should just go with #14 or #12. Even #16 lamp cord is sufficient at lower power levels.
jn
Hello,
Have to agree, worrying about speaker cables carrying up to 100khz signals is a waste of time. Speakers internal wiring, crossover networks and voice coils have enough inductance that you are never going to get anything usable out at much over 20khz in the best of circumstances.
Have to agree, worrying about speaker cables carrying up to 100khz signals is a waste of time. Speakers internal wiring, crossover networks and voice coils have enough inductance that you are never going to get anything usable out at much over 20khz in the best of circumstances.
"Don't worry about skin depth."
As I understood it, the issue with skin depth is not about current carrying capacity -- which is the focus of the naysayers -- but phase shift -- which was the point of analyses by people like Hawksford and Jung and Marsh (Pooge).
Just my observation from the cheap seats (but 'quiring minds... )
As I understood it, the issue with skin depth is not about current carrying capacity -- which is the focus of the naysayers -- but phase shift -- which was the point of analyses by people like Hawksford and Jung and Marsh (Pooge).
Just my observation from the cheap seats (but 'quiring minds... )
The hawksford analysis was rather inaccurate. (I'm being gentle).
The only phase shift skin can cause is the lowering of the inductance of the wire. For a twisted pair, that is a maximum drop of 30 nH per foot.
A twisted pair can do the proximity thing to the tune of roughly half it's per foot inductance, so 70 nH per foot. But again, at what range of frequencies?
If you need your speaker run to remain locked inductance over frequency, do the cat5e thingy multiple parallel runs.
I do not recall jung, marsh, or pooge saying much different.
Jn
What Jung and Marsh said was this:
"Before departing this consideration of the individual wire, it is interesting to consider how the transformation from a solid round conductor to a conducting shell might influence audio signals. Since such a process effectively does take place in a conductor with shallow skin depth, as we move from low to high frequencies, this suggests a possible change of phase relationships between fundamental/harmonic components. This is in fact what does happen, as the phase lag at the frequency f, referred to the surface, is one radian (57.3 degrees) at one skin depth inward. Thus, considering an audio signal's broad band of signal spectrum, phase degradation will take place when the skin depth is not an appreciably large percentage of the conductor, resulting in a loss of coherence." -- TAA/4/81 page 10
From my chair that was pretty much in accord with Hawksford (as specious as his rationale might have been to the untutored [mine] eye). And 57.3 degrees seems like a lot to me too. Admittedly most of this was not focused on speaker wire.
One other thing: It seems to me that the frequency limit of human hearing isn't the direct issue. A coherent signal in a conductor has to pass summed frequency signals, no? The 20kh signal sits on waves of lower frequency signals. The slope will be much steeper than for 20kh. That's why transient response is important; why, for instance, step (square wave) tests were advocated. The frequencies of interest here are much higher than the limit of human hearing.
Again, I'm asking, not telling. It's just that I find a lot of the talking points seem to talk unhelpfully past each other (which is why the debate drags on and on... ) Maybe Walt could chime in? (I understand that Richard Marsh has left the building)
"Before departing this consideration of the individual wire, it is interesting to consider how the transformation from a solid round conductor to a conducting shell might influence audio signals. Since such a process effectively does take place in a conductor with shallow skin depth, as we move from low to high frequencies, this suggests a possible change of phase relationships between fundamental/harmonic components. This is in fact what does happen, as the phase lag at the frequency f, referred to the surface, is one radian (57.3 degrees) at one skin depth inward. Thus, considering an audio signal's broad band of signal spectrum, phase degradation will take place when the skin depth is not an appreciably large percentage of the conductor, resulting in a loss of coherence." -- TAA/4/81 page 10
From my chair that was pretty much in accord with Hawksford (as specious as his rationale might have been to the untutored [mine] eye). And 57.3 degrees seems like a lot to me too. Admittedly most of this was not focused on speaker wire.
One other thing: It seems to me that the frequency limit of human hearing isn't the direct issue. A coherent signal in a conductor has to pass summed frequency signals, no? The 20kh signal sits on waves of lower frequency signals. The slope will be much steeper than for 20kh. That's why transient response is important; why, for instance, step (square wave) tests were advocated. The frequencies of interest here are much higher than the limit of human hearing.
Again, I'm asking, not telling. It's just that I find a lot of the talking points seem to talk unhelpfully past each other (which is why the debate drags on and on... ) Maybe Walt could chime in? (I understand that Richard Marsh has left the building)
Last edited:
You are discussing, that is a good thing. Your posts have been all good...
The standard skin depth equations are based on penetration of a planar E/M wave traveling normal to a conductive surface. So all the stuff happening comes from outside the conductive surface, a wave in free air hitting the surface.
Skin depth of the current within a current carrying conductor is a reaction the the current within.
Hawksford made the same mistake in thinking, but he even had a problem with his test setup, using a steel wire in place of copper. Yes, wire with much higher than free space permeability.
Jn
"Skin depth of the current within a current carrying conductor is a reaction the the current within."
[sigh] So don't leave us dangling here.
What are the correct skin depth equations based on signal-generated currents (and waves) -- and by the way aren't the subject EM waves generated by the signal? or where are these coming from? and why would the people who consider skin depth important for current carrying capacity not worry about eliminating them rather than fussing with litz wire? -- and what is the phase shift per depth? Different than one radian? More? less" Non-existent??
I realize that this might seem patently obvious to some here but for the rest of us, the blanket dismissals have never helped us to understand... understanding is the best first step to quashing rumour (light -- illumination -- is the best disinfectant) and it might be that something taken for granted is the obstacle to enlightenment.
There's a lot I don't know, eh. Which might explain the plethora of question marks!
[sigh] So don't leave us dangling here.
What are the correct skin depth equations based on signal-generated currents (and waves) -- and by the way aren't the subject EM waves generated by the signal? or where are these coming from? and why would the people who consider skin depth important for current carrying capacity not worry about eliminating them rather than fussing with litz wire? -- and what is the phase shift per depth? Different than one radian? More? less" Non-existent??
I realize that this might seem patently obvious to some here but for the rest of us, the blanket dismissals have never helped us to understand... understanding is the best first step to quashing rumour (light -- illumination -- is the best disinfectant) and it might be that something taken for granted is the obstacle to enlightenment.
There's a lot I don't know, eh. Which might explain the plethora of question marks!
The correct equations are called bessels. And they are not exactly easy.
What is typically used is called the "exponential approximation". At 20khz in a 1mm wire, the exponential approximation is off by about 50%.
When I get back onto my laptop (not IPad) I will look for a good picture and explanation to link.
The gist is, when the current within a wire changes, the eddy currents in the wire fight the rate of change, and they do so in a way that the current near the center is cancelled, and near the edge they are added to.
Jn
What is typically used is called the "exponential approximation". At 20khz in a 1mm wire, the exponential approximation is off by about 50%.
When I get back onto my laptop (not IPad) I will look for a good picture and explanation to link.
The gist is, when the current within a wire changes, the eddy currents in the wire fight the rate of change, and they do so in a way that the current near the center is cancelled, and near the edge they are added to.
Jn
I appreciate you doing this even if, in the end, I can't follow the maths.
I would still judge a half a radian to be significant (that said in all my glorious ignorance!)
(I may also be otherwise occupied for the rest of the day... but I am interested and open to understanding this) Thanks.
I would still judge a half a radian to be significant (that said in all my glorious ignorance!)
(I may also be otherwise occupied for the rest of the day... but I am interested and open to understanding this) Thanks.