|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.13.232.33
In Reply to: RE: "Ordinary" magnet wire posted by kurt s on October 24, 2010 at 00:15:09
“"It may be a total shock to some people to know that a typical 8-ohm four layer woofer voice coil contains about 120 feet of number 28 solid copper wire.”
I am not aware of any driver designers who would be surprised by that revelation. None I know buy into any of the more exotic wire beliefs, all attribute the magnitude of the effect of wire based on the wires individual components effect in an equivalent circuit.
“The "solid copper wire" is magnet wire that is usually very pure copper no matter what the source.”
But not for the reasons you might think, in order to make copper wire, the copper is processed after smelting (sort of a electrolytic plating process) to remove impurities like silver and gold , many of which impair the wire drawing process, soft pure copper is best.
Insulation, this too is chosen by it’s properties, magnet wire insulation has to be thin and tough, in some loudspeaker VC’s, it must stay intact at 350degrees C. For voice coils, an adhesive is usually added to the outer surface.
Old drivers never had power compression because the wire would melt off below that temperature, now we have high power voice coils where the temperature can reach or exceed 350C (the resistance of the wire has already doubled at 230C).
“The same reason the magnet wire gauge choice is made in the first place for voice coils - it's skin depth optimization for the frequencies covered and that just happens to work well (plus tighter coupling in the coils). “
Ah, no. From a driver design (engineering) view it is more like this.
The volume of the magnetic gap times density is proportional to cost.
The strength of the motor one creates is proportional to the BL / sqr root of Rdc.
The upshot is that for a given cost, the strength of the motor is optimized by minimizing the Rdc of the winding in that equation.
In the chart of conductors, copper is next to the best and the best is only a little better while being harder to work with and much more expensive.
Skin effect is a real thing, real at audio frequencies but irrelevant because the magnitude of the effect it has is insignificant. Curiously, the poorer the conductor the material is, the higher in frequency one needs to go for a given skin depth.
In drivers where one can trade a greater gap volume for lower mass, a lesser conductor like aluminum is often used. With aluminum, one has about half the conductivity but one fourth the mass so it is also used in voice coils and of course most overhead power wires.
Keep in mind, from a skin depth view a 10 ft long, 1 inch dia copper bar has a deviation that begins well within the audio band BUT if one plots it’s resistance ( impedance) vs frequency, then one sees even at 20KHz it still has a very very low resistance. Relative to the load it drives, it has an insignificant effect, this is what one would measure looking at each end of the wire and comparing the differences.
“A little extra wire resistance on mids/highs does not seem to hurt a lot when used as a speaker wire to it, when so much is already in the voice coils. Mids/tweeters might be better off with some other solution when driven directly from separate amps and electronic crossover, though. Different block diagrams have different optimum solutions for interface.”
This is true mostly. The drivers “motor” is electrically in series with the VC’s Rdc and so adding some resistance comparable to say a fraction of the Rdc, has minimal effect other than lowering the level. For a VC motor, it is CURRENT that produces force to move the radiator, it is the Voltage across the load that produces current flow.
This is still true at low frequencies BUT modern drivers are configured to use the amplifier as part of the system damping. Here a copper penny clamped across a woofers terminals replicated a very high damping factor, the drivers motor is in series with the Rdc, any motion of the motor produces a voltage which presents a current across Rdc which then produces an opposing force. If you have a woofer lying around in the open, place it on the table and tap on the radiator, then short the terminals together and try again.
Thus, for a woofer one does want a low series resistance, if one looks to see at what point there is no further change, it would be when the series R in the wire is about 1/20 to 1/40 the drivers Rdc.
That is why the amplifier damping factor wars in the 80’s were only a marketing numbers game, past about 20-50 had no effect on woofer control.
Hope that helps
Tom
Follow Ups:
that household wiring ( Romex and such) is spec'ed to be 99.9% pure. That is the equivalent to reagent grade in a chem lab.
Stu
-Kurt
Edits: 07/05/11
Hi
“I believe the simplest concept of 12 gauge Romex is fantastic in the bass but not further up in frequency.”
Here is why. The need for a low resistance source for the woofer was explained already. The current flowing in a speaker cable is automatically balanced, that is two equal but opposite currents.
A magnetic field surrounds a conductor carrying current, that field gives the wire the property of series inductance. If one places two equal but opposite conductors in close proximity, the equal but opposite magnetic fields cancel each other out.
One can see this if you wind a coil of wire and measure it’s inductance then close the opening until the two sides are close together and now the inductance is much much lower.
In a speaker cable, the conductors are not infinitely close together in fact in Romex or hardware store extension cord, the conductors are separated for safety by filler and insulation.
As a result of that spacing, there is what is called leakage flux, flux that does not couple or cancel and becomes a series L. As a result, the series inductance per foot can add up if the length is sufficient to roll off the high end. It’s parallel C and series L per foot is a good way to look at wire, a short wire is always better than a longer one of the same type.
I have a tower that I measure speakers with in an anechoic condition, I needed to run a 125 foot speaker cable to the tower and needed that cable to have negligible effect (1/10dB or less) on the speakers measurements. I had samples of a number of commercial speaker cables and many types of AC cords and RF cables.
After many measurements with an HP network analyzer, I found that a pair of cross coupled coax cables gave the best practical results. The 125 foot cable had approximately the same series L of the woven Kimber hifi cable (12 feet long) sample while having less Rdc and less C per foot.
Anyway, if you interested in trying what appears to be VERY good electrically speaking, take advantage of true wire science, try some lmr-400 ultra flex at a screaming $1.20 a foot . (I found some at a hamfest)
http://www.universal-radio.com/catalog/cable/coax.html
Coax is an attempt to minimize the series inductance to a minimum value with near perfect field cancellation. Depending on the type, some of it also has very low C per foot, all in all this area (RF) is where the vast majority of cable research has been conducted because all the secondary effects becomes ever larger as one climbs the decade after decade above audio in frequency.
Best,
Tom
-Kurt
Edits: 07/05/11
Hi
I think John was an engineer at Peavey, I remember him from AES a good while ago, not familiar with his speaker cable idea.
I am not sure you got my meaning though, ALL of the “small things” like parallel C, series L, become the limiting factors much higher in frequency or in much longer wires. If one wants “low loss” dielectric, then looking at “ultra low loss coax” is logical.
Note some of them have a mostly air dielectric using a spiral web insulator, others have foam TFE etc. Hard line, the next step up from cable is like two copper tubes one inside the other or one tube with a solid center. This can be made with even lower losses but is not flexible or logical in light of the speakers electrical load and frequency range.
Orders of magnitude more work go into making these kinds of cables because the market is vastly larger AND the effects more critical.
At a reasonable distance, cable like the 400 would appear to be blameless based on what difference one would see at one end vs the other and it’s parallel C loading vs other choices.
Audio and loudspeaker design is my hobby, my love and for much of my life part of my job.
My background has been in making things, sometimes audio transducers, sometimes other things.
I would say the most useful things to have worked on in this area was in trying to improve output transformers a large (250W) tube amp and then being Principal investigator on a research contract in high frequency electromagnetic levitation.
A high frequency is used because a material like glass or ceramic has a large skin depth and so one can induce an eddy current through most of it’s volume. Once there is a current, then the material heats and becomes more conductive and easier to levitate (that eddy current produces an opposing magnetic field)
Anyway, picture how one would make a small one turn coil out of solid copper plated with heavy silver and pass 100 to 300 amps through it at 10 to 13Mhz. Then, set up six of those orthogonally and drive them 120 degrees apart in phase and the top opposite from the bottom and one has a way to average out the force so that melted material stay nearly spherical. The transformers and networks needed to match a one turn coil to the 300W, 50 ohm amplifiers were a real “learning experience” which if anyone says that about your job to you, it means run away..
http://adsabs.harvard.edu/abs/1991AdSpR..11...79D
http://ip.com/patent/US5150272
I have developed a number of other electromagnetic / acoustic things too, some of which are transducers, speakers or audio related a few which have not issued are the basis of the company’s products I work for now.
http://www.google.com/patents?tbs=bks%3A1&tbo=1&q="Thomas+j+danley"&btnG=Search+Patents
Twisted pair is a great configuration, the thin insulation and tight twist has an effect like shielding in that it is hard to induce a differential signal with an external field. The only possible down side would be the measured parallel C value relative to the driving impedance.
On the other hand, making the cable as short as practical minimizes any effect, all the undesirable stuff is “by the foot”.
For an interconnect, one might try Teflon wire wrap wire as a way to lower the C with a slightly larger spacing and lower dielectric constant (than the magnet wire insulation).
The other old days trick at short wave frequencies (before tfe was popular) was to wind a tight twisted pair and then unwind it a bit so there was a casual airspace between most of the conductors.
Seems to me there was a thumb rule for the optimum twist per foot relative to the wire diameter but i don't remember what is was.
Best,
Tom
-Kurt
Edits: 07/05/11
-Kurt
Edits: 07/05/11
-Kurt
Edits: 07/05/11
I would find the arguments for non-linear effects to be more plausible if they could be quantified, e.g. percent distortion. To be most effective, the quantitative results should be justified by both modeling (simulation) and measurement (physical experimentation).
In the absence of any numbers my working assumption will continue to be that non-linear effects in active electronics are vastly more important than in wiring. The one exception is at connectors, where corrosion is known to create significant non-linear effects which can easily be measured and heard.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
-Kurt
Edits: 07/05/11
I have no problem with people reporting that they hear differences. My problem starts when people (often marketing people as Tom has pointed out, but presumably not in your case) start positing physical explanations for these perceptions. Often their explanations don't pass the smell test. In the case at hand, your theory hasn't even risen to a level at which the faintest odor might be detected.
I object to half-baked science, be it physics, biology, psychology or statistics. If you want to stay firmly in the realm of art, then you will get no complaints from me. But you are positing physics. There should be numbers. If there is no measurement equipment capable of the required precision there is still the possibility of simulation.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
-Kurt
Edits: 07/05/11
Sorry you are not feeling well. Best wishes for a quick recovery.
I believe in connectors causing audible distortion based on the obvious noise and distortion that I have heard and could easily have measured had I been so inclined. In most cases it was readily cured by twisting the connectors, volume control knobs and switches that were affected. Over many decades I have repeatedly observed this effect and its cure.
If you want numbers on connectors you can find them on the Internet regarding the situation with cell phone towers. There are other articles, but these require paid subscriptions so I haven't read them. Also Google "rusty bolt effect".
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Hi Tony
Boy, you said a mouthful and made a very important point.
One thing you see in audio marketing using science is that they often describe things which are real electronic / physical effects and occasionally described with some degree of accuracy as well.
What is lacking nearly always is anything which might put that effect in perspective. For example, reading a cable brochure about skin effect, one would conclude it is a major consideration in all cable or wiring.
Yet, if one were dealing with AC power distribution or were looking at the effect skin effect has at 20KHz when a typical speaker cable connects a speaker to an amplifier, one would conclude it is usually a non-issue, superseded by much larger effects like series L and R which are more configuration and cross section & conductivity related.
The reason for the two sides of this and other issues is something like that in fishing I think.
The fisherman describes a fish he is proud of, describes how it felt to fight him into the boat over what seemed like an eternity, how it made the rod bend in half, how you struggled to net him and so on. He describes an even in the physical world from the perspective of the fisherman, based on what his eyes see plus what he knows feels and expects. If you were selling things to a fisherman, it would be best to pitch the product in those same terms as the fisherman relates to, to acknowledge fishing as something like a religious experience.
If you were designing the internal workings like the gears of that reel or other part of that product your best off knowing what you’re dealing with in objective engineering terms.
Re-examining the event;
Measured reality is how long the fish is, it’s cubic volume, it’s weight, measured reality is the record of force on the line over time, it is a plot of the acceleration, velocity and vector of the fish over the 293 seconds the event lasted. Engineering requires all the data related to the physics of the operation and awareness of the ideal goals.
So we have hifi which runs the entire spectrum from the real to the imaginary, where each person draws the line is governed by what you believe to be true, what you believe to be true are things which you have seen, heard of or been demonstrated.
Curiously, or not surprisingly depending how you view it, a faction in hifi is also conditioned to be paranoid / hostile about even trying to see if you can hear the difference between A and B without knowing which was which. Funny stuff, not many places left where inquiry is frowned on and for sure, it's not where significant progress is likely.
Fwiw, our earliest solid state semiconductors (diodes) were the oxidized electrical contact between two materials. A crude radio could be built using a razor blade and graphite pencil lead and tank coil, with the contact between the oxidized razor blade and graphite rectifying the radio signals and then fed to very sensitive earphones.. The old finned selenium rectifiers being the end of that oxidized boundary approach, switching to bulk conductors in solid materials like germanium and then silicon which lead to the IC and modern solid state.
In RF, many cables have silver plated connectors, silver oxidizes but it is not a semi conductor and has low resistance which is easily displaced.
Best,
Tom
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: