Home |
Search |
Today's Posts |
#41
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Per Stromgren wrote in message
line. No, I was not aware of the Smith chart, I'm afraid. But it seems to me (after some 30 minutes of Googling I just performed) that it is just a calculation tool, and it does not per se introduce any alternative theory. As far as I can see, there is no mention of this tool to work in the application you just showed us. Most sources say "RF broadband use only" or something to that effect. Or did I miss something? Per. A Smith will show the change of impedance of a transmission line, for a line as short as 0.002 WL. This is not a new theory, it's just the way it is. If one rotates the cursor of a Smith chart by 0.002 WL, one will be able to read, how the impedance of a 0.002 WL line changes. There is no low frequency limit to Smith Charts. The only limits a the transmission line must be at least 0.002 WL long and the line must have very low series resistance. The case I cited was a 100 ft line, of 12 gage wire. A line of 100 Ft is just barely long enough to allow Smith chart calculations. A 12 gage cable has just barely low enough series resistance, to allow the Smith chart to work accurately. Of one thing I'm su calculating the loss of a 100 ft line, using discrete component models, will not give correct results. Bob Stanton |
#42
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#43
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#44
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#47
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
In article ,
(Bob-Stanton) wrote: (Dick Pierce) wrote in message No, it shows nothing of the kind. It shows that you have a model of a physical phenomenon that gives an answer that satisfies you. Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless. The *Smith chart* is not my model, it's Bell Lab's model. It is a transmission line solving graph, that has been in use for more than 65 years. It's taught by every electrical engineering college in the world. It's in every RF text book. But, if you think it is worthless, that is your right. You could become quite well know by proving that, as a predictor of transmission line behavior, it is worthless! Find out how a transmission line's behaviour is different at, say, 100 Hz than it is at, say, 10 MHz. Then decide if a Smith chart will tell you the "truth" at 100 Hz. Hint: The "high frequencyapproximations" for Zo and so on DO NOT WORK at 100 Hz. Isaac |
#48
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#49
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#50
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Dick Pierce) wrote in message No, it shows nothing of the kind. It shows that you have a model of a physical phenomenon that gives an answer that satisfies you. Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless. The *Smith chart* is not my model, it's Bell Lab's model. No, YOU are the one, Mr. Stanton, who is championing it's use in an application for which it was not designed. It is therefore YOUR model. YOU made it that way, thus it's up to YOU to prove its efficiacy for the use YOU are claiming. To date, you have not provided a single shred of measured data that supports your claim. It is a transmission line solving graph, that has been in use for more than 65 years. It's taught by every electrical engineering college in the world. It's in every RF text book. Yes, you are right, it's in every RF text book. Why isn't it in every electrical engineering text book that does NOT deal with RF? But, if you think it is worthless, that is your right. Mr. Stanton, your dishonest misrepresentations are showing again. Please do not lie about what I said. Here is EXACTLY what I said: "Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless." Now, Mr. Stanton, that's very different tha what you're implying, isn't it. Since YOU haven't shown us ONE SINGLE SHRED OF EMPIRICAL EVIDENCE to show that YOUR use of a Smith chart is appropriate for an application for which it was not design, YOUR use stands completely unsubstantiated and, until YOU come up with the supporting evidence, it's no more useful than any other model which has NO supporting evidence. You could become quite well know by proving that, as a predictor of transmission line behavior, it is worthless! You are becoming well known in your own right of using models without ever taking ANY effort of seeing whether they work or not. Mr. Stanton, who uses 100 foot long speaker cables in their home? What sort of idiotic nonsense is that? I don't think anyone disagrees that there are significant transmission line effects at 10 Hz with speaker cables stretched from here to the moon. But so what? Your model fails on several fronts: you have utterly failed to show that your models predict actual physical behavior better than other models, you have selected a completely unrealistic, pathological case that, by your own admission "barely fits" the model. |
#51
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Dick Pierce) wrote in message No, it shows nothing of the kind. It shows that you have a model of a physical phenomenon that gives an answer that satisfies you. Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless. The *Smith chart* is not my model, it's Bell Lab's model. No, YOU are the one, Mr. Stanton, who is championing it's use in an application for which it was not designed. It is therefore YOUR model. YOU made it that way, thus it's up to YOU to prove its efficiacy for the use YOU are claiming. To date, you have not provided a single shred of measured data that supports your claim. It is a transmission line solving graph, that has been in use for more than 65 years. It's taught by every electrical engineering college in the world. It's in every RF text book. Yes, you are right, it's in every RF text book. Why isn't it in every electrical engineering text book that does NOT deal with RF? But, if you think it is worthless, that is your right. Mr. Stanton, your dishonest misrepresentations are showing again. Please do not lie about what I said. Here is EXACTLY what I said: "Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless." Now, Mr. Stanton, that's very different tha what you're implying, isn't it. Since YOU haven't shown us ONE SINGLE SHRED OF EMPIRICAL EVIDENCE to show that YOUR use of a Smith chart is appropriate for an application for which it was not design, YOUR use stands completely unsubstantiated and, until YOU come up with the supporting evidence, it's no more useful than any other model which has NO supporting evidence. You could become quite well know by proving that, as a predictor of transmission line behavior, it is worthless! You are becoming well known in your own right of using models without ever taking ANY effort of seeing whether they work or not. Mr. Stanton, who uses 100 foot long speaker cables in their home? What sort of idiotic nonsense is that? I don't think anyone disagrees that there are significant transmission line effects at 10 Hz with speaker cables stretched from here to the moon. But so what? Your model fails on several fronts: you have utterly failed to show that your models predict actual physical behavior better than other models, you have selected a completely unrealistic, pathological case that, by your own admission "barely fits" the model. |
#52
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Isaac Wingfield wrote in message news:isw-
Find out how a transmission line's behaviour is different at, say, 100 Hz than it is at, say, 10 MHz. Then decide if a Smith chart will tell you the "truth" at 100 Hz. Hint: The "high frequencyapproximations" for Zo and so on DO NOT WORK at 100 Hz. I know that the formula for characteristic impedance, Zo = sqrt(L/C), is only a high frequency approximation. The correct formula for all frequencies, is: Zo = sqrt((R + jwL)/(G + jwC)). A Smith Chart will give accurate results if: jwL (which is also XL) is much larger than R (cable series resistance). A 12 gage cable has about 0.2uH/ft inductance and 0.0016 Ohms/ft series resistance. At 20 KHz, the XL/ft is 0.251 Ohms. That is 15.6 times larger than the R/ft. Data published by Belden and others, gives insertion loss figures for 100 ft of 12 gage, two wire, speaker cable, terminated by 4 Ohms. Analysis of these data leads me to believe that the Smith Chart still gives accurate results when the XL is 15.6 times larger than R. At 100 Hz, the XL is *much* smaller than R. At 100 Hz, the speaker cable is simply a resistance. Smith Chart results wouldn't apply. Bob Stanton |
#53
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Isaac Wingfield wrote in message news:isw-
Find out how a transmission line's behaviour is different at, say, 100 Hz than it is at, say, 10 MHz. Then decide if a Smith chart will tell you the "truth" at 100 Hz. Hint: The "high frequencyapproximations" for Zo and so on DO NOT WORK at 100 Hz. I know that the formula for characteristic impedance, Zo = sqrt(L/C), is only a high frequency approximation. The correct formula for all frequencies, is: Zo = sqrt((R + jwL)/(G + jwC)). A Smith Chart will give accurate results if: jwL (which is also XL) is much larger than R (cable series resistance). A 12 gage cable has about 0.2uH/ft inductance and 0.0016 Ohms/ft series resistance. At 20 KHz, the XL/ft is 0.251 Ohms. That is 15.6 times larger than the R/ft. Data published by Belden and others, gives insertion loss figures for 100 ft of 12 gage, two wire, speaker cable, terminated by 4 Ohms. Analysis of these data leads me to believe that the Smith Chart still gives accurate results when the XL is 15.6 times larger than R. At 100 Hz, the XL is *much* smaller than R. At 100 Hz, the speaker cable is simply a resistance. Smith Chart results wouldn't apply. Bob Stanton |
#54
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Bob-Stanton) wrote in message One full rotation around the chart equals one wavelength. Let me correct the statement (above) that I wrote in the previous message. One full rotation around the Smith Chart actually represents *1/2 wavelength* of transmission line. Mr. Stanton, how about some other corrections? Try the following on for size: 1. "You know, with all my hand waving about the Smith chart, I have never once actually measured a cable to see if I am right or not. I have three potential answers, one from my beloved Smith chart, one from some RF simulation program of unknown pedigree, and one from a lumped parameter analysis also of some unknown pedigree. What's REALLY missing is the actual data to see if ANY of the predictions are right." 2. "I have a hypothesis. It's counter to other hypotheses floating around out there. In order to see whether my hyptheses is any more valuable that others, I have to make falsifiable predictions. To be falsifiable, my predictions have to be unambiguously confirmable by measurements. And if the error in the measurements is greater that the difference between my predictions and those of others, nothing has been proven. "I haven't done any of that yet. And, until I do so, my model has no real value. Things that have no real value are without worth. Things that have no worth are, well, 'worthless.'" 3. "Actually, using 100 foot speaker cables was kinda silly. I redid the calculations using 10 feet and, you know, you just can't use a Smith chart any more. See, a wavelength in a cable at 20 kHz, even considering a 50% velocity factor, is 7500 meters which is about 24,750 feet or some 4.69 MILES. Now, a speaker cable 10 feet long will have a delay on the order of 0.07 degrees between one end and the other at 20 kHz. Now, at the highest rate of change, 0.07 degrees will result in a difference in voltage between one end and the other to the tune of 0.00122: due to the transmission delay alone, that means if I put 10 volts in at 20 kHz, what will come out the other end will be 9.988 volts at the other. That's a difference of 0.012 volts due to the transmission line delay. "That would mean, it seems, that the ENTIRE cables is acting pretty much as a single element. Why, with only a MAXIMUM difference in voltage of about one part in a thousand between one end and the other, it sure seems to be ACTING like a single part. I mean, golly, for it to be a transmission line, doesn't there have to be a significant difference in time and thus phase (since time and phase are duals of one another, mathematiclly) in order for the very concept of 'transmission' to have sufficient strength to dominate the behavior of the system we hold before us? "So I says to myself, 'Self, how does that compare to simple Ohmic attenuative effects?' So using my 10 feet of 12 gauge annealed coper wire, which has an Ohmic loss of 0.00158 Ohms per foot, I found out that the total series resistance is on the order of 0.032 Ohms which, considering a nominal load impedance of 8 ohms, results in a voltage across the load of 9.96 volts. "'My goodness,' I hear myself exclaim, 'the effects of just ONE of the lumped-parameter effects is at least as great as the alledged (but, to this point UTTERLY unverfied by ANY corrobarative experimental data) transmission line effects! "Could I be wrong? Could I have overemphasised the important of the transmission line effects? Is it possible that I really DO have to throw out my collection of antique Smith charts, like my wife has been telling me, because after 23 attempts to unload them on EBay wity a reserve price of $186,240, not once did I get a bid? "Zut alors! What am I to do?" Try those corrections for a start. |
#55
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Bob-Stanton) wrote in message One full rotation around the chart equals one wavelength. Let me correct the statement (above) that I wrote in the previous message. One full rotation around the Smith Chart actually represents *1/2 wavelength* of transmission line. Mr. Stanton, how about some other corrections? Try the following on for size: 1. "You know, with all my hand waving about the Smith chart, I have never once actually measured a cable to see if I am right or not. I have three potential answers, one from my beloved Smith chart, one from some RF simulation program of unknown pedigree, and one from a lumped parameter analysis also of some unknown pedigree. What's REALLY missing is the actual data to see if ANY of the predictions are right." 2. "I have a hypothesis. It's counter to other hypotheses floating around out there. In order to see whether my hyptheses is any more valuable that others, I have to make falsifiable predictions. To be falsifiable, my predictions have to be unambiguously confirmable by measurements. And if the error in the measurements is greater that the difference between my predictions and those of others, nothing has been proven. "I haven't done any of that yet. And, until I do so, my model has no real value. Things that have no real value are without worth. Things that have no worth are, well, 'worthless.'" 3. "Actually, using 100 foot speaker cables was kinda silly. I redid the calculations using 10 feet and, you know, you just can't use a Smith chart any more. See, a wavelength in a cable at 20 kHz, even considering a 50% velocity factor, is 7500 meters which is about 24,750 feet or some 4.69 MILES. Now, a speaker cable 10 feet long will have a delay on the order of 0.07 degrees between one end and the other at 20 kHz. Now, at the highest rate of change, 0.07 degrees will result in a difference in voltage between one end and the other to the tune of 0.00122: due to the transmission delay alone, that means if I put 10 volts in at 20 kHz, what will come out the other end will be 9.988 volts at the other. That's a difference of 0.012 volts due to the transmission line delay. "That would mean, it seems, that the ENTIRE cables is acting pretty much as a single element. Why, with only a MAXIMUM difference in voltage of about one part in a thousand between one end and the other, it sure seems to be ACTING like a single part. I mean, golly, for it to be a transmission line, doesn't there have to be a significant difference in time and thus phase (since time and phase are duals of one another, mathematiclly) in order for the very concept of 'transmission' to have sufficient strength to dominate the behavior of the system we hold before us? "So I says to myself, 'Self, how does that compare to simple Ohmic attenuative effects?' So using my 10 feet of 12 gauge annealed coper wire, which has an Ohmic loss of 0.00158 Ohms per foot, I found out that the total series resistance is on the order of 0.032 Ohms which, considering a nominal load impedance of 8 ohms, results in a voltage across the load of 9.96 volts. "'My goodness,' I hear myself exclaim, 'the effects of just ONE of the lumped-parameter effects is at least as great as the alledged (but, to this point UTTERLY unverfied by ANY corrobarative experimental data) transmission line effects! "Could I be wrong? Could I have overemphasised the important of the transmission line effects? Is it possible that I really DO have to throw out my collection of antique Smith charts, like my wife has been telling me, because after 23 attempts to unload them on EBay wity a reserve price of $186,240, not once did I get a bid? "Zut alors! What am I to do?" Try those corrections for a start. |
#56
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Bob-Stanton" wrote in message m... snip Bob Stanton Oh dear, lots of heat, not much light ;-) The Smith Chart was introduced as a convenient visual method to avoid some tedious math. It is the equivalent of a slide rule. The original question was "How to measure speaker cable inductance and capacitance?" Get hold of an LCR meter. Short (connect together live and ground at the far end of the cable), select "Inductance" on the LCR meter, and measure the inductance (to be picky, try and use an LCR meter that does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). Select "Resistance"on the meter. Measure DC resistance. Open circuit the far end of the speaker cable, select "Capacitance" on the LCR meter, measure capacitance. Now, work out the loss at 20kHz from a low impedance source of the (lumped element) network you have measured into an 8/4 ohm (whatever) load. Pretty low, huh? That says the "lumped" approximation is just fine. Incidentally, this is sometimes used by _very_ reputable RF cable manufacturers to check the characteristic impedances of their cables (the extra measurement needed is with a ruler). Regards Ian |
#57
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Bob-Stanton" wrote in message m... snip Bob Stanton Oh dear, lots of heat, not much light ;-) The Smith Chart was introduced as a convenient visual method to avoid some tedious math. It is the equivalent of a slide rule. The original question was "How to measure speaker cable inductance and capacitance?" Get hold of an LCR meter. Short (connect together live and ground at the far end of the cable), select "Inductance" on the LCR meter, and measure the inductance (to be picky, try and use an LCR meter that does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). Select "Resistance"on the meter. Measure DC resistance. Open circuit the far end of the speaker cable, select "Capacitance" on the LCR meter, measure capacitance. Now, work out the loss at 20kHz from a low impedance source of the (lumped element) network you have measured into an 8/4 ohm (whatever) load. Pretty low, huh? That says the "lumped" approximation is just fine. Incidentally, this is sometimes used by _very_ reputable RF cable manufacturers to check the characteristic impedances of their cables (the extra measurement needed is with a ruler). Regards Ian |
#58
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#59
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#60
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Per Stromgren wrote in message . ..
On 10 Nov 2003 17:21:37 -0800, (Bob-Stanton) wrote: Well, being in "every RF text book" tells us something, does it not? Yes. Radio frequency techniques can not automatically be applied to the snail pace of audio frequencies. Yes. If you still think that the transmission line model can be applied to AF, please show us some reference that supports you. Because all sources I (and I suppose many others here) have seen indicates the opposite to what you say. Your agrument is called: appeal to authority. Sometimes, if someone has no logical way of proving his point, he cites an authortiy. What you should do is, tell us *why* the authority you cited said what he said. For someone with a hammer, everything looks like nails! (I prefer a screw gun myself.) :-) Just because someone argues only one side of an issue, does not automatically mean he is unobjective. Bob Stanton |
#61
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Per Stromgren wrote in message . ..
On 10 Nov 2003 17:21:37 -0800, (Bob-Stanton) wrote: Well, being in "every RF text book" tells us something, does it not? Yes. Radio frequency techniques can not automatically be applied to the snail pace of audio frequencies. Yes. If you still think that the transmission line model can be applied to AF, please show us some reference that supports you. Because all sources I (and I suppose many others here) have seen indicates the opposite to what you say. Your agrument is called: appeal to authority. Sometimes, if someone has no logical way of proving his point, he cites an authortiy. What you should do is, tell us *why* the authority you cited said what he said. For someone with a hammer, everything looks like nails! (I prefer a screw gun myself.) :-) Just because someone argues only one side of an issue, does not automatically mean he is unobjective. Bob Stanton |
#62
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#63
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#64
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Dick Pierce) wrote in message No, YOU are the one, Mr. Stanton, who is championing it's use in an application for which it was not designed. It is therefore YOUR model. YOU made it that way, thus it's up to YOU to prove its efficiacy for the use YOU are claiming. To date, you have not provided a single shred of measured data that supports your claim. That right, I don't have any measured data. Perhaps you could grab some cable from your junk box, and grace us with some measurements. Mr. STanton, it's not my job or anyone else's to do YOUR work for you. YOU are the one making the claim, YOU are the one who has utterly failed to to provide ANY evidence for your claim. Yes, you are right, it's in every RF text book. Why isn't it in every electrical engineering text book that does NOT deal with RF? The Smith Chart is for solving transmission line problems. Only RF and video people deal with transmission lines on a regular basis. So, why would other areas of electrical engineering care about Smith Charts? Because, by your own admission, they don't apply. Mr. Stanton, your dishonest misrepresentations are showing again. Please do not lie about what I said. Here is EXACTLY what I said: "Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless." Dick, you also said: "It shows you have a model of a physical phenomenon that gives an answer that satisfies you." (You forgot that part.) I interpeted your words to mean you doubted the accuracy of the model I used (the Smith Chart). Your interpretation is either wrong or dishonest. But, now that we both agree the Smith Chart is accurate, let us go on. No, once again, you are being out and out dishonest. I never agreed with anything at all. And, until YOU provide evidence supporting your claim, YOU are unable to go on. I do model a lot of stuff. No, I don't run tests to see whether my models work. (I'll leave that up to you Dick.) Wrong, it's up to YOU. Since I have almost 100% comfidence in my models, why should I test them? Since the Roman Catholic Church had 100% confidence in their model of the universe with the earth at the center and all else moving around it in harmonic spheres, why should HEY test them? Mr. Stanton, who uses 100 foot long speaker cables in their home? Bill Gates? Wow, talk about itotic appeal to authority! I don't think anyone disagrees that there are significant transmission line effects at 10 Hz with speaker cables stretched from here to the moon. But so what? I disagree. A transmission line that long would have so much resistive loss, that it wouldn't in any way function as a "well behaved" transmission line. It is a common misconception that a long, long, speaker line would have "transmission line" characteristics, at audio frequencies. So, let's summarize your extraordinary claims, which seem to have gelled into two major points: 1. Very short speaker wires, like 10 feet, MUCH shorter than a wavelength, DO behave like tyransmission lines 2. Very long speaker wires MUCH longer than a wavelength, DO NOT behave like transmission lines. Your model fails on several fronts: you have utterly failed to show that your models predict actual physical behavior better than other models, you have selected a completely unrealistic, pathological case that, by your own admission "barely fits" the model. Oh really! Please show me mathematically where my theory fails. Thank you in advance. As you, thus far, have completely failed to show where your theory DOES work, there's no work for ANYONE else top do. You have admitted that at 100 feet, your theory "barely" works. You have admitted you have NO evidence to support your theory. You have admitted that you have total confidence in your theory irrespective of your complete lack of supporting data. Gee, golly, Mr. Stnton, it seems you have left NOTHING left for anyone else to do on your theory. |
#65
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
(Bob-Stanton) wrote in message . com...
(Dick Pierce) wrote in message No, YOU are the one, Mr. Stanton, who is championing it's use in an application for which it was not designed. It is therefore YOUR model. YOU made it that way, thus it's up to YOU to prove its efficiacy for the use YOU are claiming. To date, you have not provided a single shred of measured data that supports your claim. That right, I don't have any measured data. Perhaps you could grab some cable from your junk box, and grace us with some measurements. Mr. STanton, it's not my job or anyone else's to do YOUR work for you. YOU are the one making the claim, YOU are the one who has utterly failed to to provide ANY evidence for your claim. Yes, you are right, it's in every RF text book. Why isn't it in every electrical engineering text book that does NOT deal with RF? The Smith Chart is for solving transmission line problems. Only RF and video people deal with transmission lines on a regular basis. So, why would other areas of electrical engineering care about Smith Charts? Because, by your own admission, they don't apply. Mr. Stanton, your dishonest misrepresentations are showing again. Please do not lie about what I said. Here is EXACTLY what I said: "Until such time as you check those answers against the real physical behavior in a way that is clear, unambiguous, consistent and predictive, your models are worthless." Dick, you also said: "It shows you have a model of a physical phenomenon that gives an answer that satisfies you." (You forgot that part.) I interpeted your words to mean you doubted the accuracy of the model I used (the Smith Chart). Your interpretation is either wrong or dishonest. But, now that we both agree the Smith Chart is accurate, let us go on. No, once again, you are being out and out dishonest. I never agreed with anything at all. And, until YOU provide evidence supporting your claim, YOU are unable to go on. I do model a lot of stuff. No, I don't run tests to see whether my models work. (I'll leave that up to you Dick.) Wrong, it's up to YOU. Since I have almost 100% comfidence in my models, why should I test them? Since the Roman Catholic Church had 100% confidence in their model of the universe with the earth at the center and all else moving around it in harmonic spheres, why should HEY test them? Mr. Stanton, who uses 100 foot long speaker cables in their home? Bill Gates? Wow, talk about itotic appeal to authority! I don't think anyone disagrees that there are significant transmission line effects at 10 Hz with speaker cables stretched from here to the moon. But so what? I disagree. A transmission line that long would have so much resistive loss, that it wouldn't in any way function as a "well behaved" transmission line. It is a common misconception that a long, long, speaker line would have "transmission line" characteristics, at audio frequencies. So, let's summarize your extraordinary claims, which seem to have gelled into two major points: 1. Very short speaker wires, like 10 feet, MUCH shorter than a wavelength, DO behave like tyransmission lines 2. Very long speaker wires MUCH longer than a wavelength, DO NOT behave like transmission lines. Your model fails on several fronts: you have utterly failed to show that your models predict actual physical behavior better than other models, you have selected a completely unrealistic, pathological case that, by your own admission "barely fits" the model. Oh really! Please show me mathematically where my theory fails. Thank you in advance. As you, thus far, have completely failed to show where your theory DOES work, there's no work for ANYONE else top do. You have admitted that at 100 feet, your theory "barely" works. You have admitted you have NO evidence to support your theory. You have admitted that you have total confidence in your theory irrespective of your complete lack of supporting data. Gee, golly, Mr. Stnton, it seems you have left NOTHING left for anyone else to do on your theory. |
#66
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#67
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#68
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Ian" wrote in message ...
"Bob-Stanton" wrote in message m... snip Bob Stanton Oh dear, lots of heat, not much light ;-) Did you want light (enlightment)? You are the first person who has asked. :-) The Smith Chart was introduced as a convenient visual method to avoid some tedious math. It is the equivalent of a slide rule. Of course the Smith Chart is not as accurate as doing the math. I actually did do the math, but that would be too tedious and hard to explain on this forum. So I thought, I'll use the Smith Chart! It's a relativly simple, graphical way to show how a transmission line works. The original question was "How to measure speaker cable inductance and capacitance?" Get hold of an LCR meter. Short (connect together live and ground at the far end of the cable), select "Inductance" on the LCR meter, and measure the inductance (to be picky, try and use an LCR meter that does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). My guess is: "1571" was Waynes' lucky number? Select "Resistance"on the meter. Measure DC resistance. Open circuit the far end of the speaker cable, select "Capacitance" on the LCR meter, measure capacitance. Now, work out the loss at 20kHz from a low impedance source of the (lumped element) network you have measured into an 8/4 ohm (whatever) load. OK lets do it! I did a Google search and found Belden, and BRTB Canada, specified loss for 100 ft of 12 gage speaker cable. Belden specs stated 11% or 0.5 dB loss for 140 ft of Brillance cable. That seems a little too low. BRTB Canada gave 12.21% loss, with a 4 Ohm load, for 100 ft, 12 gage, premium speaker cable. I'll use the lossier BRTB specification: 12.21% = 0.6 dB loss. The Inductance of 12 gage is 0.2uH/ft or 20 uH for 100 ft. BRTB spec for capacitance was 27.4 pF/ft (2740 pF for 100 ft). R= 0.002 Ohms/ft Now lets make a discrete component model: --------20 uH ------0.2 Ohms------------ | | | Source 2740 pf 4 Ohms | | | ----------------------------------------- What did you get for loss, at 20 Khz? I got -1.77 dB. Pretty low, huh? That says the "lumped" approximation is just fine. No. I'd say not low enough. BRTB specs said 0.6 dB loss. Now lets do a (crude) transmission line model: ----- 0.2 Ohms-------| 100ft, 86 Ohm, t-line |--- | | Source 4 Ohms | | ------------------------------------------------- I got -0.7 dB. The transmission line model's results are quite a bit closer to the manufactures specification. Not only did the lumped element model not work very accurately, but the transmission line model (which some say doesn't work at 20 KHz) gave close results. Incidentally, this is sometimes used by _very_ reputable RF cable manufacturers to check the characteristic impedances of their cables (the extra measurement needed is with a ruler). RF cable manufactures use either a return loss bridge, or a network analyzer to measure RF cable impedance (return loss). Regards, Bob Stanton |
#69
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Ian" wrote in message ...
"Bob-Stanton" wrote in message m... snip Bob Stanton Oh dear, lots of heat, not much light ;-) Did you want light (enlightment)? You are the first person who has asked. :-) The Smith Chart was introduced as a convenient visual method to avoid some tedious math. It is the equivalent of a slide rule. Of course the Smith Chart is not as accurate as doing the math. I actually did do the math, but that would be too tedious and hard to explain on this forum. So I thought, I'll use the Smith Chart! It's a relativly simple, graphical way to show how a transmission line works. The original question was "How to measure speaker cable inductance and capacitance?" Get hold of an LCR meter. Short (connect together live and ground at the far end of the cable), select "Inductance" on the LCR meter, and measure the inductance (to be picky, try and use an LCR meter that does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). My guess is: "1571" was Waynes' lucky number? Select "Resistance"on the meter. Measure DC resistance. Open circuit the far end of the speaker cable, select "Capacitance" on the LCR meter, measure capacitance. Now, work out the loss at 20kHz from a low impedance source of the (lumped element) network you have measured into an 8/4 ohm (whatever) load. OK lets do it! I did a Google search and found Belden, and BRTB Canada, specified loss for 100 ft of 12 gage speaker cable. Belden specs stated 11% or 0.5 dB loss for 140 ft of Brillance cable. That seems a little too low. BRTB Canada gave 12.21% loss, with a 4 Ohm load, for 100 ft, 12 gage, premium speaker cable. I'll use the lossier BRTB specification: 12.21% = 0.6 dB loss. The Inductance of 12 gage is 0.2uH/ft or 20 uH for 100 ft. BRTB spec for capacitance was 27.4 pF/ft (2740 pF for 100 ft). R= 0.002 Ohms/ft Now lets make a discrete component model: --------20 uH ------0.2 Ohms------------ | | | Source 2740 pf 4 Ohms | | | ----------------------------------------- What did you get for loss, at 20 Khz? I got -1.77 dB. Pretty low, huh? That says the "lumped" approximation is just fine. No. I'd say not low enough. BRTB specs said 0.6 dB loss. Now lets do a (crude) transmission line model: ----- 0.2 Ohms-------| 100ft, 86 Ohm, t-line |--- | | Source 4 Ohms | | ------------------------------------------------- I got -0.7 dB. The transmission line model's results are quite a bit closer to the manufactures specification. Not only did the lumped element model not work very accurately, but the transmission line model (which some say doesn't work at 20 KHz) gave close results. Incidentally, this is sometimes used by _very_ reputable RF cable manufacturers to check the characteristic impedances of their cables (the extra measurement needed is with a ruler). RF cable manufactures use either a return loss bridge, or a network analyzer to measure RF cable impedance (return loss). Regards, Bob Stanton |
#71
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#72
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Dick Pierce" wrote in message om... (Bob-Stanton) wrote in message . com... "Ian" wrote in message ... "Bob-Stanton" wrote in message m... snip Bob Stanton snip does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). My guess is: "1571" was Waynes' lucky number? Yes, that's a guess, and it would be wrong. Hint, divide 1571 by 2 pi and see what it gives you. Try to figure out why one would do that with a frequency. Now, why am I not surprised at these two answers? Regards Ian |
#73
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
"Dick Pierce" wrote in message om... (Bob-Stanton) wrote in message . com... "Ian" wrote in message ... "Bob-Stanton" wrote in message m... snip Bob Stanton snip does the test in the audio range - ISTR the Wayne Kerr boxes used something like 1571Hz (extra credit if you can work out why ;-)). My guess is: "1571" was Waynes' lucky number? Yes, that's a guess, and it would be wrong. Hint, divide 1571 by 2 pi and see what it gives you. Try to figure out why one would do that with a frequency. Now, why am I not surprised at these two answers? Regards Ian |
#74
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
|
#76
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Bob-Stanton wrote:
...snips.. Somehow you guys seem to keep poking sticks in each other's cages. I don't see that any of this hand waving will get you out of this poking match. First I'd like to pose the question: with the thousands of components and feet of wire that audio goes through between the recorded source and the playback system (including room acoustic effects) why are people so passionate about the few feet of speaker cable? Let me try to explain this again. Any two wire line will be a "transmission line" if the resistance/ft is much lower than the XL/ft. You suggested (wrongly) that a long line, at 10 Hz, would be a transmission line. The XL(inductive reactance) of 12 gage cable is only 0.00001 Ohms/ft at 10 Hz. That XL is much lower than the resistance/ft. The line will not be a transmission line, at 10 Hz, no matter what it's length. Ummm, why? What's the magic about the reactance to resistance ratio that makes "any two wire line" a transmission line? Seems to me you only made a statement rather than an explanation. 2. Very long speaker wires MUCH longer than a wavelength, DO NOT behave like transmission lines. That right, if the resistance/ft is too high, it will be only a "line". [ see "Ummm" above... ] You have admitted that at 100 feet, your theory "barely" works. I didn't say a 100 ft line "barely" works. I said basically, that 20 KHz is probably the lower frequency limit for "transmission line" behavior. And we're still only talking about fractions of a dB difference ...at high frequencies. Heck, opening the curtains or another person in the room likely has more impact on the sound... How does anyone know what may have been heard when recording and mixing the sound... [ returning to lurking mode ] Ron Capik -- |
#77
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Bob-Stanton wrote:
...snips.. Somehow you guys seem to keep poking sticks in each other's cages. I don't see that any of this hand waving will get you out of this poking match. First I'd like to pose the question: with the thousands of components and feet of wire that audio goes through between the recorded source and the playback system (including room acoustic effects) why are people so passionate about the few feet of speaker cable? Let me try to explain this again. Any two wire line will be a "transmission line" if the resistance/ft is much lower than the XL/ft. You suggested (wrongly) that a long line, at 10 Hz, would be a transmission line. The XL(inductive reactance) of 12 gage cable is only 0.00001 Ohms/ft at 10 Hz. That XL is much lower than the resistance/ft. The line will not be a transmission line, at 10 Hz, no matter what it's length. Ummm, why? What's the magic about the reactance to resistance ratio that makes "any two wire line" a transmission line? Seems to me you only made a statement rather than an explanation. 2. Very long speaker wires MUCH longer than a wavelength, DO NOT behave like transmission lines. That right, if the resistance/ft is too high, it will be only a "line". [ see "Ummm" above... ] You have admitted that at 100 feet, your theory "barely" works. I didn't say a 100 ft line "barely" works. I said basically, that 20 KHz is probably the lower frequency limit for "transmission line" behavior. And we're still only talking about fractions of a dB difference ...at high frequencies. Heck, opening the curtains or another person in the room likely has more impact on the sound... How does anyone know what may have been heard when recording and mixing the sound... [ returning to lurking mode ] Ron Capik -- |
#78
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
On Thu, 13 Nov 2003 18:05:47 GMT, Ron Capik wrote:
Bob-Stanton wrote: ...snips.. Somehow you guys seem to keep poking sticks in each other's cages. I don't see that any of this hand waving will get you out of this poking match. First I'd like to pose the question: with the thousands of components and feet of wire that audio goes through between the recorded source and the playback system (including room acoustic effects) why are people so passionate about the few feet of speaker cable? It is something that is visible. These are the same idiots who'll spend $1000 on a 2' cable while ignoring all the other connections internal to their equipment, or who'll spend $400 on a power cord to connect to 2 miles of plain power wiring. |
#79
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
On Thu, 13 Nov 2003 18:05:47 GMT, Ron Capik wrote:
Bob-Stanton wrote: ...snips.. Somehow you guys seem to keep poking sticks in each other's cages. I don't see that any of this hand waving will get you out of this poking match. First I'd like to pose the question: with the thousands of components and feet of wire that audio goes through between the recorded source and the playback system (including room acoustic effects) why are people so passionate about the few feet of speaker cable? It is something that is visible. These are the same idiots who'll spend $1000 on a 2' cable while ignoring all the other connections internal to their equipment, or who'll spend $400 on a power cord to connect to 2 miles of plain power wiring. |
#80
|
|||
|
|||
How to measure speaker cable inductance and capacitance?
Per Stromgren wrote in message
And what does it tell you? It tells me that the Smith Chart may be a clever way to solve a computational *RF* problem, at least when we had no better alternatives than slide rules and log books. Perhaps there are better ways nowadays, who knows? The Smith Chart is based on standard transmission line theory. Like a sliderule it has limited accuracy. A sliderule may tell you that 2 times 3 is: 5.99, but we know it is based on sound priciples. The Smith Chart is slightly inaccurate also. But, just like as the sliderule, its fundamentals are valid. Radio frequency techniques can not automatically be applied to the snail pace of audio frequencies. Yes. Right. So why do you suggest we do use this method, then? Because you say so? I'm saying a transmission line model of a two-wire line, will give slightly more accurate results than a discrete component model. OK. What do you want me to do in order to convince you? Well, runing a physical test would be nice. Or you could prove from basic theory, that small wavelength lines don't "work". Citing sources won't do, abvoiusly. Is there any experiment result that would show you that you are wrong? As you know, our dialog is classical. You say A. I say: A is not what a lot of their people say. You say: prove that the others are right. Why should I? Isn't you the one who ought to prove "the others" wrong? I don't think they are wrong, I think you may be misunderstanding what they said. You may be confusing textbook statements about transmission lines at *low frequencies*, with transmission lines that have *small wavelenghts*. They are not the same thing at all! Your turn. (I think I know the next step.) I have presented a theory based on well known facts and formulas. What more do I need to do? Of course being a totally objective person, and being a totally honest person, and being ever *so* humble as well, I don't insist that everyone agree that I am right. (Even if I am.) Have an nice day. Bob Stanton |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Speaker cables, your opinions appreciated | High End Audio | |||
rec.audio.car FAQ (Part 4/5) | Car Audio | |||
cabling explained | Car Audio | |||
Measure Speaker Cable! | High End Audio | |||
Making my own speaker cables... | High End Audio |