Thank you Jim. Al is known for his knowledge and willingness to help others. He is also much better in explaining things. Al, have you ever been a teacher?
80 responses Add your response
Al (almarg) and kijanki, I would trust your "expectation bias" over some others so called facts any day. Thank you both for your contributions to this audio forum. I never stop learning from what the two of you have to say. Sometimes I may not fully understand, but eventually it sinks in. Thanks again, Jim |
Thanks once more, Kijanki. Your post just above is excellent IMO, as usual. As far as the audibility of dielectric absorption in audio cables is concerned, my reading of various technical and anecdotal references to that effect I've seen over the years suggests to me that it stands a good chance of being audibly significant in many applications. However I have never seen either an analysis or measured data that would provide a quantitative perspective on it, in the context of audio cables. So FWIW my own "expectation bias" is in the direction of that effect being great enough in degree to be an audibly significant contributor to cable differences, in many systems. But as far as I am aware information doesn’t seem to be available that would provide insight that is more specific. Best regards, -- Al |
Geofkaitt, Yes, it is about 66% = 5ns/m (I inverted it wrong). Teflon is probably a little faster. They also use foam Teflon to lower dielectric constant. In addition wires can be in the hollow tubes since dielectric constant of the air is pretty much the same as vacuum (approx. 1) Jea48, Dielectric can possibly cause distortion since it absorbs and releases energy. Example of this is capacitor that is charged, discharged by shorting and then opened. Voltage on the capacitor will start growing back from zero to many volts. It is very pronounced in electrolytic capacitors. It happens because dielectric stores and returns energy. It is called Dielectric Absorption and is also related to Permittivity. Returning voltage when signal has already different level can cause distortion. Looking at the lines of audio cable I can say that price is proportional to dielectric used. PVC, that you mentioned is pretty bad while Polypropylene is better, Teflon better yet and oversized tubes of foam Teflon are the best (Acoustic Zen Absolute IC). How much of this is audible I don't know. AZ Absolute sounded "cleaner" to me (more refined), but it can be placebo effect (I expected it). Dielectric Constant also affects capacitance between wires. Capacitance of typical cable is around 25pF/ft while AZ Absolute IC is 6pF/ft. Can capacitance of IC be a source of tiny distortion? Al, what is your opinion on dielectric absorption in audio cables? Can this be audible? |
kijankikijanki, Can the type of dielectric used cause distortion of an analog signal as it travels through an interconnect? What frequencies, would you say, are affected the most? Example PVC vs Teflon? Tough question.... Can it be measured? Jim |
Thank you, Kijanki. I was just about to post that numerous references can be found on the web indicating that the dielectric constant of Teflon is in the vicinity of 2.0, or even a bit more, not 1.0. Also, the 70-85% figure Geoff cited is of course at best an average or typical propagation velocity, and examples of audio cables having propagation velocities that are significantly slower and significantly faster are easily found. Steve (Williewonka), the book you referenced looks like an excellent read! I note, btw, that the section your link goes to was authored by Bill Whitlock, of Jensen Transformers, who like Ralph Morrison is a noted authority on such matters. And I note that Mr. Morrison himself is referred to in Mr. Whitlock’s writeup. Also, if I may be a bit presumptuous, let me extend kudos for your interest in gaining as thorough a technical understanding of such matters as possible, to complement what I know is your very extensive practical experience experimenting with various cable configurations. Regards, -- Al |
Typical velocity factors, numbers on left are percentages of velocity of light in a vacuum. Transmission line 95–99 Open-wire "Ladder" Line 83 RG-6 Belden 1189A coaxial cable 82 RG-8X Belden 9258 coaxial cable (foamed polyethylene dielectric) 80 Belden 9085 twin-lead 66 Belden 8723 twin shielded twisted pair stranded (polypropylene insulator)[10] 66 RG-213 CXP213 coaxial cable (solid polyethylene dielectric) |
Dielectric constant of Teflon is about 2. Vacuum has dielectric constant of 1. Yes, Teflon will slow down electromagnetic wave. Insulator will slow down electromagnetic wave by amount based on its ability to store energy - Permittivity. Dielectric constant is just relative Permittivity. This speed of electromagnetic wave thru typical insulated wire is about 60% of the speed of light in the vacuum. For typical cable it comes to about 5ns/m and it is exactly true for cat5 cable. There is no different electromagnetic wave for audio signals and other signals. |
As far as the directionality thread is concerned Al and Atmasphere have definitely not proved their points. Furthermore, it should be pointed out much of Al’s argument (as is often the case) is an Appeal to Authority, citing experts to support his argument. Even citing his own expertise, not to mention Atmasphere’s. That’s an appeal to authority. You know, a logical fallacy. Geez, all you would have to do to win any (rpt any) technical argument is say well, I found this guy so and so and he says such and such so I must be right. Cheers |
almargAl, For the latter. I am glad you answered Geoff’s post though. Jim |
Let’s try a different approach. Teflon is obviously a very good dielectric material, right? It has a dielectric constant of approximately 1.0 if I’m not mistaken. Which means that electomagnetic waves will not (rpt not) be slowed significantly through Teflon. On the other hand, we know that the audio signal - which (I think there is agreement on this) is an electromagnetic wave - is found to travel only around 70-85% of the velocity of light in a vacuum. I believe this means that the electromagnetic wave must be traveling through the copper, not the dielectric. His else could you explain the discrepancy? |
Kijanki, thanks for your characteristically excellent technical input. Williewonka 8-24-2017 Hi Steve, First, be sure to keep in mind, as you no doubt realize, that a voltage must always be defined with respect to some reference. Given that, in the example you cite above the neutral conductor would indeed be at or very close to zero volts, **relative to the circuit grounds/signal grounds of the two components.** And probably in most (but not all) designs relative to AC safety ground and earth ground as well. But those facts do not have any inconsistency with what I said in my earlier posts. Consider the simple example of a 120 volt light bulb. When it is turned on via the switch on the wall, if you were to individually measure the current in the "hot wire" and the "neutral wire" that are connected to it you would measure exactly the same amount of current in both. Even though the neutral wire is at or very close to zero volts relative to earth ground and to AC safety ground. I’ll take a look at the book you referenced later today or tonight. Best regards, -- Al |
Al - I just found this book and it is making things a lot clearer for me Audio Engineering Explained https://books.google.ca/books?id=aLDpAwAAQBAJ&pg=PT224&lpg=PT224&dq=is+the+neutral+condu... Let me know what you think please Cheers - Steve |
Jim (Jea48), Jc4659, and Kijanki, thanks very much for your kind words. Jim (Jea48), I’m not sure if your most recent post is suggesting that I try to explain why Geoff’s comment is incorrect, or that I refrain from doing so to avoid having this heretofore constructive thread go downhill the way the recent thread on wire directionality has. But I’ll assume the former, perhaps incorrectly. Geoffkait 8-24-2017 Geoff, to be precise, skin effect means that as frequency progressively increases above a certain frequency (which depends on the diameter of the conductor), "current density" (with "current" defined as in one of my earlier posts) decreases to a greater degree at progressively greater depths. That causes a progressive increase in the resistance of the conductor at progressively higher frequencies. "Audio frequencies travel inside the conductor," to use your words, in the sense that the movement of electrons, at the very slow drift velocity I referred to, is a very small back and forth oscillatory motion occurring at the same frequency or frequencies for which energy is being conveyed in the electromagnetic wave. As the +/- polarity of the applied voltage changes, at a given frequency, the direction of that slow movement of electrons changes correspondingly. (And actually, to be precise, I should say "net movement of electrons," because random movement of some electrons is always occurring to some degree). In the recent wire directionality thread which you participated in extensively, Jim (Jea48) quoted a statement by Ralph Morrison, a world renowned authority on such matters, and the author of several textbooks, which contradicts your assertion that the electromagnetic wave travels within metallic conductors. That assertion was also contradicted in another thread here by a noted designer of highly respected world class audio electronics, as well as by me and several other technically knowledgeable posters. During the course of my lengthy career and schooling in electrical engineering I have never seen such an assertion ever made by anyone other than yourself. If you can cite a seemingly credible reference supporting your contention I will attempt to explain why it is either incorrect or is being misinterpreted. Regards, -- Al |
Geoffkait, electric charge still travels back and forth inside of the wire between terminals hence skin effect still applies. It starts for the copper at 20kHz at about gauge 18. Al was only explaining difference between electric charge travelling back and forth (electric current) and energy delivery from source to load in form of electromagnetic wave outside of the cable. Load has voltage drop between terminals hence we have electric field while current in the wire produces magnetic filed. Interaction of these fields causes power draw by the load. Source also has electric and electromagnetic fields that cause power output. As Al stated this energy is delivered in one direction only in form of electromagnetic wave outside of the cable - from source to load. Direction is determined by Poynting Vector. Electric and magnetic fields are perpendicular to each other and determine direction of Poynting vector (direction of energy transfer). It is also true for DC. |
Al - thanks for the informative response However, I need your help again... Let’s take the IC connecting two components as an example... - the signal conductor has an AC signal on it - the neutral conductor is connected to the neutral sides of each component On well designed components the neutral side of the circuit should always be at zero vaults - especially if grounded If both components are well designed, then the neutral sides of the their respective circuits would be at zero volts, Therefore, the neutral conductor of the IC should also be at zero volts - yes? What exactly am I missing here ? Sorry for being a pain and thanks in advance |
| Post removed |
If the audio signal travels through the *dielectric* and not (rpt not) through the metal conductor I suppose we can throw out the whole skin effect idea, which says most audio frequencies travel *inside* the metal conductor at some depth, with only very high frequencies, perhaps above "audio frequencies," traveling near the surface, I.e., skin. How can audio frequencies travel inside the conductor when the audio signal - the electromagnetic wave - travels outside the conductor? |
almarg Al, Great post! Jim . |
The entire "theory" of the "neutral" is that it, (Shall return the unbalanced load) to the system. And yes it's conductive properties need to be "At least" as functional in resistance as it's positive counterpart. If it is not? And feel free to try this at home! Plug in a lamp with just the "hot" wire. The bulb in the lamp will typically explode. Because there wasn't enough of a path to ground and no neutral. So what happens when you use a lesser cable for your neutral? Unless your speaker is 100% percent efficient. "AND" can use the entire load, It "can" become distortion if there is too much resistance. Of course this should only happen at higher volumes with most systems. I hope this helped. |
Thanks, Jerry! Steve, when I refer to currents in the two conductors that are equal except that they are moving in opposite directions (i.e., current in one conductor is moving toward the load when current in the other conductor is moving toward the source), another way to look at it, that amounts to saying the same thing but may make it more clear, is that the currents in both conductors are moving in the same direction but around a loop. The loop consisting of the two conductors plus the input circuit of the load plus the output circuit of the source. And between each half-cycle the direction the current is traveling around that loop reverses. Best regards, -- Al |
Hi Steve, You raise good questions, which get into some complexities that are not obvious. "The signal," and the energy that it conveys, is conducted through neither of the conductors. It is conducted in the form of an electromagnetic wave, which propagates at a substantial fraction of the speed of light in a vacuum, and propagates through the dielectric which surrounds the conductors. The exact propagation speed is dependent primarily on what is known as the "dielectric constant" of the particular insulation. Putting aside reflection effects that can occur mainly at RF frequencies as a result of impedance mismatches, and assuming that the load is essentially resistive, that energy propagates in just one direction, from the source of the signal to the load. However, that propagation of the signal and its energy is intimately related to movement of electrons within both of the conductors, which takes place in both directions (the direction alternating in each of the two conductors, assuming we’re not dealing with DC), and which takes place at an ***extremely*** slow velocity that is referred to as "drift velocity." In the case of electrical signals that are conducted via wires (as opposed, for example, to being radiated through the air or a vacuum), the extremely slow movement of electrons within the conductors and the near light speed movement of the signal and its energy are intimately related, as I said, and one would not occur without the other. A way to visualize it is that at the instant a signal voltage is applied to the source end of a cable, a **very** slow movement of electrons will occur into one of the two conductors at that end of the cable, and out of the other of the two conductors at that end of the cable, corresponding to the +/- polarity of the signal at that instant. At the other end of the cable, and at all points in between, there will be a similar slow movement of **different** electrons, with the response of those electrons being delayed from the response of the electrons at the source end of the cable by the amount of time it takes "the signal" to traverse the corresponding cable length (at near light speed). What can be referred to as "the current," as opposed to "the signal," can be considered as corresponding to the number of electrons traversing a given cross-section of a conductor in a given amount of time. One ampere of current, for example, corresponds to one coulomb per second, where one coulomb corresponds to the amount of charge possessed by about 6.2 x 10^18 electrons. So assuming that only two paths exist between the source and the load, namely the two conductors in a single cable, "the current" being conducted by both conductors in response to an applied signal is in fact identical, except that when it is moving in one direction in one conductor it is moving in the other direction in the other conductor. And in the case of audio signals, or any kind of signal other than DC, the directions in the two conductors alternate between each half-cycle of the waveform. So with the slight possible exception I mentioned earlier about RFI/EMI pickup, in the case of a speaker cable the two conductors are of equal importance. In the case of a line-level analog interconnect, on the other hand, IMO the "ground" or "return" conductor should if anything be considered to be **more** important than the "signal" or "hot" conductor. The reason being that the characteristics of the return conductor may affect susceptibility to ground loop-related high frequency noise or low frequency hum, depending on the internal grounding configuration and other aspects of the designs of the particular components that are being connected. So given the foregoing it hopefully becomes clear that your statement that... ... when I think about speaker cables, the "energy" in the signal conductor must be very different from the neutral side simply because by the time the signal gets through the speaker voice coil, most of it has been converted into the movement of the driver, so the neutral must be quite different - doesn’t it?... is not a correct statement because the transfer of energy to the load goes hand-in-hand with current (movement of charge carriers, i.e., electrons) in **both** conductors. With that movement being equally important in the two conductors, and (putting aside the possible ground loop and RFI/EMI effects I’ve mentioned) being identical in the two conductors aside from being in opposite directions at any instant of time. Hopefully that clarifies more than it confuses :-) Best regards, -- Al |
@almarg - you raise a point that has puzzled me for a very long time After all, they are conducting the same current, just in opposite directions at any given instantGranted, the current may be flowing in the opposite direction in the signal conductor, but can the same be said of the neutral? Is the neutral actually the opposite of of the signal? When you consider - the neutral conductor in IC's are actually connected to the -ve side of the circuit(s) in the connected components and it's only the signal conductor actually "carrying" the alternating signal, things start to look a little different from the "return path" approach most people are familiar with. Granted - you do have to have to connect both signal and neutral to both components to "complete the circuit", but is the neutral actually transferring energy that reflects the signal, i.e. except for its opposite polarity? All of my cables now use different quality conductors, mainly because I have tried using the same conductor for signal and neutral and found it provided no discernible advantage. Having said that the quality of the neutral conductor I use is quite high EXAMPLE: in my IC's I use a solid silver signal conductor and a quality copper neutral conductor of approximately twice the gauge of the silver. Using a copper signal & copper neutral results in a less dynamic sound than the Silver Signal - copper neutral IC, But using a silver neutral with a silver signal provided no benefit over the silver signal and copper neutral IC. Also, when I think about speaker cables, the "energy" in the signal conductor must be very different from the neutral side simply because by the time the signal gets through the speaker voice coil, most of it has been converted into the movement of the driver, so the neutral must be quite different - doesn't it? As I said - it has perplexed me for a very long time - even more so since I experimented with cables and experienced my observations. Any help understanding this would be appreciated Cheers - Steve |
Good comments by Steve (Williewonka). I would add that if the + and - conductors referred to in the OP are not in close proximity, and preferably twisted together in some manner, the inductance of the cable will be considerably increased. If the impedance of the speakers is low at high frequencies (as it is, for example, in the case of many electrostatics), and/or if the cable length is long (inductance is proportional to length, for a given cable type), that may result in perceptible rolloff of the upper treble, and dull or sluggish sounding transients. Also, in terms of wire quality I would consider the negative conductor to be no less important than the positive conductor. After all, they are conducting the same current, just in opposite directions at any given instant. A conceivable exception to that, however, is that the amplifier might be more susceptible to RFI/EMI picked up by the cable and introduced into its feedback loop (if it has one) from the positive conductor than from the negative conductor. Regards, -- Al |
@chalmersiv - although you will still hear music the sound quality between the two different combinations you mention may be different If you step up to a higher quality ground wire then the quality will not be impacted I currently use a high quality 16 gauge silver plated Mil-spec wire for the signal and a reasonable quality 13 gauge (2 x 16 gauge) for the neutral. The sound is extremely good, but that is due in large part to the geometry of the cable rather than the quality of the conductors used. See: http://image99.net/blog/files/d048bbacfce9bcad4a025be804771d9a-76.html I have experimented with different combinations of conductor and having a thicker gauge neutral seems to provide a better sounding cable. Although it would be nice to get away with cable from Lowes, there is no substitute for a conductor made from a reasonably high quality copper I have also experimented with Romex, 12 gauge Extension cord (from Home Depot), Kimber Kable and the wire in the link above, using various geometries. The cables in the link above provides the best sound quality to date, surpassing all of the various store bought cables I have tried to date Hope that helps - Steve |
