Magnet Wire Question
Al, I agree that skin effect might be unimportant. I only stated the boundary since cable companies (AQ FAQ) claim it is important and make all sorts of expensive arrangements like flat woven tape or helical twist of multiple wires on round hollow core (my Acoustic Zen Satori). It is far fetched concept but they know more than I do. There were sound differences between cables I used that I couldn't explain. Tighter twist doesn't hurt and makes often handling of the wire easier. My router happens to be 5GHz. Such frequencies, once enter the box can either find LC circuit to couple to, or some nonlinear element to mix on - even while "searching" for ground return path. In addition to mixing there is also rectification phenomenon where many decades lower bandwidth amplifier converts high frequency signal into very small level DC (because of uneven positive and negative slew rates) that becomes "audible" when offending signal is amplitude or frequency modulated. We're talking microscopic levels - but why even to allow this garbage to enter amplifier's box. Wires inside of my Rowland model 102 box have very tight twist. Why audiophiles select thick gauges? I'd like to know. Perhaps to reduce inductance? |
Al, I agree that skin effect might be unimportant. I only stated the boundary since cable companies (AQ FAQ) claim it is important and make all sorts of expensive arrangements like flat woven tape or helical twist of multiple wires on round hollow core (my Acoustic Zen Satori). It is far fetched concept but they know more than I do. There were sound differences between cables I used that I couldn't explain. Tighter twist doesn't hurt and makes often handling of the wire easier. My router happens to be 5GHz. Such frequencies, once enter the box can either find LC circuit to couple to, or some nonlinear element to mix on - even while "searching" for ground return path. In addition to mixing there is also rectification phenomenon where many decades lower bandwidth amplifier converts high frequency signal into very small level DC (because of uneven positive and negative slew rates) that becomes "audible" when offending signal is amplitude or frequency modulated. We're talking microscopic levels - but why even to allow this garbage to enter amplifier's box. Wires inside of my Rowland model 102 box have very tight twist: http://www.gzhifi.com/uploads/userup/0809/052315454K9.jpg Why audiophiles select thick gauges? I'd like to know. Perhaps to reduce inductance? |
This is one of the very rare occasions on which I must disagree somewhat with my learned A'gon colleague Kijanki. Theoretical skin effect in copper at 20kHz starts at gauge 18.16 and 14 gauge will therefore experience a very slight rise in resistance at 20kHz, and in the ultrasonic region, compared to their resistance at lower frequencies. But that increased resistance will still be a tiny fraction of the impedance of this and most other dynamic (as opposed to electrostatic) speakers at those frequencies, and therefore completely inconsequential. Cable marketing literature notwithstanding. Use drill - you'll never make even twisting by hand. Put at least one turn per inch. Twisted wire effectively reduces noise pickup for frequencies that have wavelength longer than pitch of the twist.The suggestion I made of approximately 3 turns per foot corresponds to 4 inches per turn. For noise frequencies and other electromagnetic radiation that may be propagating through the air, a 4 inch wavelength corresponds to a frequency of 3 gigaHertz. The corresponding figure for noise frequencies that may be propagating in the cable itself will be on the order of 2 gigaHertz. Given that both the speakers and the amplifier have bandwidths that are many thousands of times lower than those frequencies, extending those cutoffs higher by making the turns closer together would seem unlikely to accomplish anything. Damping factor doesn't matter.As I said earlier, "it could be argued that 31 is more than adequate for this and most speakers." And, yes, it may not matter. But then why do so many audiophiles settle on cables that are 12 or 10 or 8 or even lower gauges, often at considerable added expense compared to narrower gauges? Thicker wire would reduce inductance (only a little) but you already do that by twisting. Twisting reduces inductance but also increases capacitance. It should not make a difference with speaker wire.Agreed. Best regards, -- Al |
Theoretical skin effect in copper at 20kHz starts at gauge 18. Use drill - you'll never make even twisting by hand. Put at least one turn per inch. Twisted wire effectively reduces noise pickup for frequencies that have wavelength longer than pitch of the twist. Damping factor doesn't matter. 4 ohm speaker's impedance is mostly resistive and is most likely around 3 ohms. This resistance is in series. Xover inductor in series with the woofer is about 0.1 ohm. Your effective DF is less than 1.3. As long as your amp and wires are 10x lower (0.3 ohm) it will change DF only by 10%. 0.128 ohm would be perfectly fine. Thicker wire would reduce inductance (only a little) but you already do that by twisting. Twisting reduces inductance but also increases capacitance. It should not make a difference with speaker wire. Make few versions with very tight, tight and loose twisting and listen. magnet wire is cheap. Good luck. |
I couldn't find an impedance curve on the CP6311, but I suspect that its impedance in the top octave either rises to relatively high values, or at least doesn't decrease significantly below the nominal 8 ohm spec. So inductance is probably not all that critical. I'd suggest simply twisting the two conductors together by hand, with three or so twists per foot, and not messing with a drill. I wouldn't be surprised, though, if the impedance dropped to the vicinity of 4 ohms or so at some frequencies in the bass or mid-bass region, which would tend to increase the importance of having adequate gauge. The resistance of the 20 foot combined length of the two conductors, if they were 18 gauge, would be 0.128 ohms, as can be calculated from this wire gauge table. For frequencies at which the speaker impedance may be in the vicinity of 4 ohms that would reduce your damping factor to 4/0.128 = 31, even if the damping factor of the amplifier were infinite. It could be argued that 31 is more than adequate for this and most speakers, but my instinct would be to increase that somewhat, by going to either 16 or 14 gauge. Those gauges would increase the corresponding damping factor (assuming the amplifier's damping factor is much higher, as it would be for most solid state amps but not for most tube amps) to 50 and 79 respectively. Enjoy! Regards, -- Al |
Rax, thanks, they are enamel coated, thankfully. I read about spacing runs in a thread- may have bi wiring based comment. Al, the speakers are Usher CP6311. (87 sens and a nominal 8 ohm load) If thicker gauge would be better, I'll buy it. What gauge do you recommend to insure I give the magnet wire a fair shot? Also, I heard a suggestion to use an electric drill to twist the wires. Is that the best way you know of? Thanks guys, I appreciate the help. I tried some solid core in a woven jacket and it opened the sound up over the stranded I had in place. Curious what the minimal coated magnet wire will sound like. I tend to prefer lean over thick sound. Best- Gary |
It probably won't behave in as neutral/accurate a manner as a heavier gauge, but you may or may not find the resulting sonics to be to your liking. Some folks even prefer the sonics of 20 to 24 gauge solid core speaker cables, although IMO such gauges are likely to alter the sound considerably when used in conjunction with most speakers, unless their length is very short. The lower the impedance of your speakers, the greater the departure from neutrality/accuracy is likely to be. Assuming the wires are insulated, twisting them together, say a few turns per foot, will reduce inductance and thereby minimize any upper treble rolloff that might otherwise occur. The lower the impedance of your speakers is at high frequencies, the more important that will be. Regards, -- Al |
