Cable design, category speakers... Hmmm, some background..
The only thing a cable is supposed to do is transfer energy to the load. Any storage of energy within the cable results in the load energy being slightly lagging.. This storage is both inductive, and capacitive, and both mechanisms are lagging in nature, meaning eventually what is stored in the cable gets to the load, as the source is very low impedance...
So, to reduce this lagging storage, minimize both L and C. Unfortunately, physics gets in the way, and L and C have an inviolate relationship: That being L * C = 1034 * EDC...EDC is the effective dielectric constant, and is equal to the DC only for a coaxial structure composed of two cylindrical conductors to eliminate the inner conductor internal inductance of 15 nH per foot. For all other geometries, the effective DC can only be larger than the insulation DC..Ribbon geometries will approach this for very large ribbons with aspect rations greater than about 10. (aspect ratio being width/thickness)
The prop speed is directly related to the EDC, being V = C/sqr(EDC). (note that I neglect the term of magnetic permeability, that would make the denominator sqr(EDC * Mu). Measurement of the inductance already lumps that mu term into the EDC...If I called it "effective DC times effective Mu, it would just add to confusion since nobody(?) uses magnetic speaker wires)
So, minimize L and C the best you can, and the storage lagging mechanism reduces, and the prop speed goes up. And, as luck would have it, minimizing the cable storage happens exactly when the cable impedance is the same value as the load.
Skin effect:
Skin effect causes the storage of energy within the cable to change with frequency. At low frequencies, there is an inductance of 15 nH per foot within the conductor, which is added to the external inductance of the geometry of the two wires. As the freq goes up, this inductance goes down.
Skin effect is worse for solid conductors, and is less for stranded conductors..this is due to the fact that strand to strand conduction is less than that of a solid. When the radial conductivity of the conductor is compromised, skinning is LESS..so, stranded wires will have a little more inductance than solid, but we're talking about at most, 15 nH per foot..this is only 18% of the total for a #12 wire pair.
The conductor shape will affect capacitance and inductance, shape should be used to minimize both..coaxial being the best, ribbons second, multiple pairs third, proximity last.
The cable which has the least effect on the soundstage, with respect to adding "something", will have an effective dielectric of 1, a characteristic impedance equal to that of the load, and will have sufficient guage to keep the damping factor high. All this is quite easy to do...the question really is...will people like the sound, or does every system have different needs..
But honestly, it's all in the R, L, C, and Q of the cable. Unfortunately, there's been so much mis-information spread around that high end audio guys end up guessing and trial and error, without much in the way of science..and, realistic measurement of matched Z cables is impossible for most wire vendors, as inductance measurements at the tens of nanohenry level are very difficult to do correctly.
Cheers, John
PS..pics and graphs woulda helped, but this ain't no diy.com. |
Jneutron: What do you do about skin effect, conductor shape, EM field ( and therefore the impedance ) consistencies at various power levels, etc ??? I'd like to see what your "simple math" approach to cable design comes up with when all of the factors are taken into consideration. It is good to see that you recognize power transfer characteristics as being "important" though : ) Sean
> |
Audioengr: ""Propagation speed????? You must be joking. This is not a common-clock digital system. This is analog audio for kripes sake! Series inductance is the most important parameter, followed by capacitance and dielectric absorption.""
Yes..prop speed is important, but, not the entity itself..I explain:
Prop speed, V = 1/ sqr(LC), and also:
V = (lightspeed) / sqr(mu * epsilon)
And: L * C = 1034 * effective DC...l in nH, C in pf.
Measure L and C...you did, at 33 nH and 118 pf...your effective DC is (33*118)/1034...or, 3.76.
The prop speed is proportional to 1/sqr(effective DC).
So, the prop speed is definitely related to L: L being related to mu, wire diameter, and wire geometry....and C: C being related to epsilon, spacing, geometry.
I do agree that the term "prop speed" is rather confusing, as most seem to think it means that the transit time from the amp to the load is of any consideration...it isnt. But, the term prop speed is directly related to the DC, L, and C.
Audioengr:""If you need a low-inductance, low capacitance cable, Which you do, this is a hard combo to come by.""
Actually, it isn't hard. But, physics sets limits. I can easily make a cable which has .033 uH per foot, and about 35 pf per foot while keeping guage to about #12, but I'd prefer to keep the impedance at 8 ohms. That would be L of about .008 uH/ft..C of 135 pf/ft, diameter of about .4 inch with insulation 20 mils....or .6 inch dia with 40 mil insulation. It depends on your flashover rating requirement.
Cheers, John |
Alpha Core 7 gauge solid silver |
hey Robert.. I had and was happy with Bearlab cables for about 1.5 years, and upgraded to some Purist Audio Venustas (both IC and speaker), definitely was a worthwhile improvement. Since getting teh Purist cables, I haven't thought about cables at all.. definitely very happy with the impact they've made on my system.
btw, would love to see pics of your system. How are the cellos? |
Hi Dennis,
Ran into this topic of yours just now.
Guess life has taken you other directions in the mean time, but for what it's worth:
MIT is your best choice for Electrostats as far as I have found or heard from anyone
I have traded my Logans for Cello Grand Masters (!) last year (impressive picture of my living room if you're interested) and have thereby (had to) change the cables to static XLO's, because I need double tri-wire per side now. (this is positively unaffordable with MIT)
I guess you know about MIT 4 bi-wired cables and have probably solved your problem over the past 2 or 3 years.
I have a pair of MIT Terminator 4 bi-wires laying around and waiting for a destination if you'd be interested.
Please let me know if I can bright up your day with either a nice picture or a set of MIT4 biwires
Robert |
Dennis, I'll have to measure them - I do not recall what the specs actually are, since it's been more than 10 years since I first came up with this cable design (Silver Thunder). At that time no one ever asked about these two issues.
If it's really important for people to know, I can take the time out to do some quick measurements and post them...
Let me know?
PS. Sean, this ain't RF :- ) |
Dennis, I'll have to measure them - I do not recall what the specs actually are, since it's been more than 10 years since I first came up with this cable design (Silver Thunder). At that time no one ever asked about these two parameters.
If it's really important for people to know, I can take the time out to do some quick measurements and post them...
Let me know?
PS. Sean, this ain't RF :- ) |
Dennis, it seems you had a similar experience to mine with the Bearlabs/Stators combination... But then, Bear uses stators too doesn't he? Strangely, in my present dynamic/dipole speaker configuration the full Bearlabs set has two competitors: S-Lightning + Valhalla and S-Lighting + Kharma (the next to top). My little system can't tell which is best; my banker can and I remain, with the Bearlabs (or is it, "I remain, clueless" ;)).
Cheers |
UPDATE: I've done a lot of mulling over cables, and wasn't really sure what to do. I almost "rented" some cables from the Cable Co. www.fatwyre.com , but by the time I auditioned the cables I wanted to audition, and then paid FULL price for the cables I ultimately decided on, I decided against it (see their policy to see how they work).
From this thread, I started a private email conversation with Bear at BearLabs, and ordered some of his silver ICs and speaker cables. I just got them today... and they are AWESOME. I went from "Acoustic Research" ICs and MIT4 biwired cables to his Silver Lightning and Thunder, and my system has opened up. The music is much more enjoyable at lower levels. It sounds like it's been released from some kind of constraint. Or, like some kind of cover or haze has been removed. Much clearer, which really effects the lower and background instruments and voices. The music sounds freed...
I'm very very happy. The whole package (two sets of ICs, two sets of speaker cables) cost me somewhere under $1300. What a deal!
Just thought I'd share my findings on Bear's cables. The improvement was obvious. And that is a big deal for me, since I hardly heard the difference when I upgraded my CD player. |
Bear, how bout the specs on your cables? you said they were both low capacitance and low inductance; can you give us the measurement? |
Bear: Velocity of propagation of cables IS quite pertinent in RF applications if you want to do things "right". Using cables cut to random lengths without factoring in the velocity factor can introduce errant vswr readings and alter the load impedance that the transmitter sees. This can be used to ones' advantage though if you are trying to use the cable as an impedance transformer.
Audioengr: The specs for your cables ( Empirical Audio ) are very similar to those of Kimber 8TC. The specs i've seen on these show .03 mH / 100 pF per foot. Their nominal impedance is 17 ohms. Electrically speaking, your cables and the Kimber 8TC's should load up relatively identically ( for all practical purposes ). Sean
> |
Propagation speed????? You must be joking. This is not a common-clock digital system. This is analog audio for kripes sake! Series inductance is the most important parameter, followed by capacitance and dielectric absorption.
If you need a low-inductance, low capacitance cable, Which you do, this is a hard combo to come by. Most manufacturers dont even publish their numbers because they are so abbysmal. Here are the numbers on my cable for comparison (I consider this low inductance, moderate capacitance) I have quite a few customers with electrostatics - single-wired too:
At 10 kHz: L = .033 UH/foot, C = 118 pF/foot.
Nordost Valhalla comes close to this - measurements were published in Stereophile. |
I have mentioned in my initial post here, that I am the importer of HMS cables and NOT a private user.
Do you post in every post that you too are not a private user? I think once per thread! is sufficient. My profile also states that I am the importer of HMS. |
Tekunda, You should change your info, if you are not a private user, and represent a manufacturer! |
Whoa...Tekunda, whoa! The velocity of propagation can be argued to be entirely irrelevant to a signal that is restricted to a maximum rise time LESS than a 20kHz. sine wave (at that!)...In fact it is virtually irrelevant to most RF applications.
As far as an adjustable "current rise time selector" someone will have to tell me just what that is in reality. Since the voltage and current are directly related and linked, I see no way to adjust the "current rise time" without altering the frequency response of the cable - ie. a filter.
By definition, ALL bass speakers are "slower" than tweeters, IF you mean that the time between a min and max excursion is longer, which it HAS to be because it is a lower frequency.
But if you are talking about acceleration, that's a different bird. It may be that some woofers are "faster" than some tweeters if you are talking about acceleration.
The *real* effect of speaker cables comes mostly as it effects the waveform produced by the amplifier, especially with reactive loads and/or amps that use a lot of feedback (which most do) and/or may not be totally stable into some loads or impedances.
If you do something via a cable that adds subjective "speed" or "impact" (etc.) I'd bet you can *see* it by looking at the effect the cable has on the output of the amp when you use a nice clean square wave as a test signal - in most cases you'll see the leading edge alter and the overshoot change. I'll give you odds on that one. |
Incidentally, I have tried bearlabs speaker cable on esl with very good results. Note that I use these cables on my regular system too, so it's a reflection of taste.. Also take into account that the ESL in question has an erratic impendance curve (+1 ohm to around 20 ohm). I used a 10ft pair (IF that makes any difference).
For the record, this cable was tonally balanced & reasonably FAST -- by which I mean that transients, following one another, were subjectively perceived as appearing in quicker succession (good point, Detlof & Rives!) -- than with another cable. |
When I speak of a fast cable, I have the signal propagation speed in mind. One of our direct competitors advertises that his top cables have the fastest propagation speed in the world, but according to the calculations Dr. Strassner did, the HMS Gran Finale is even faster.
I do not say that his alone will translate in a superb cable, but I always felt that any speaker, especially a ESL, can benefit from this, especially since Mr. Strassner found out that the slower the propagation speed of the signal is, the more the signal will become distorted.
Since we discuss here a certain technical aspect, I hope I am allowed to say that one of the main features of the Gran Finale speaker cables is the ability to adjust the current rise time of the signal. It can play a role in hybrid ESL speakers, but even more so in 3 or 4 way speakers. The bass driver is always slower to react than the midrange and treble, so again, this mismatch in reaction time of the drivers to the current can translate into distortion. With the adjustable current rise time selector, you can match the bass driver to the midrange and treble drivers, gaining a less distorted sound here. |
Detlof: I had the same question--does "fast" translate to low capacitance or some other measurable characteristic, or is it a technical term like "warm, dark, dry" (I'm not trying to give anyone a hard time--I posted on the audiophile terms forum). |
Tekunda, what actually do you mean by the term faster? Do you mean dyamics and if yes, in what range of the audible spectrum ? |
Since I own electrostatic speakers myself I will dare giving it a shot, although my opinion could be biased, since I have eventually become importer of said cables.
But when I was on my cable quest, looking for the perfect set of cables and discovered that HMS Gran Finale will do a great job and at that time at least, I did not have any finical interest in mind, when I decided that in my system, with Prodigy speakers and Bel canto amps, (later I switched to Wolcott tube amps) the Gran Finale were offering a very sweet and neutral sound.
When you have ESL speakers you need a fast cable, and according to Dr. Strassner who did the math, the Gran Finale is faster than our direct competition. |
Bear, i kinda figured that they were relatively high capacitance based on the physical description and geometry that AudioEngr previously provided, but i was not sure.
As to posting a plug for your own cables, i don't see a problem with that. That is, so long as you make it clear that you are associated with the product and / or have some type of financial interest in that product line. You did that, so i don't see why anyone would get upset about it. It would be different if you went the "sockpuppet" route and raved about your product and then responded as Bear to further plug your product line. Sean
> |
I, of course am biased to my own... but Dunlavy's cables are very capacitive, that is how they achieve a nominal "8 ohm" characteristic impedance.
The cables I use and make are both low inductance and low capacitance. I use ESLs and so do some of my friends, they like the cables too.
Can I say that here? Hope so.
- |
I've gotten good results with Homegrown Audio Silver Lace speaker cables with my Acoustat 1+1 ESL's. |
Hmmm..... They recommend a speaker cable that doesn't exist. Finding a cable with all three of those characteristics at the same time might be a good trick. Nordost's are "reasonable" in terms of inductance and capacitance but have a higher nominal impedance. Kimber's have low inductance, a "medium" nominal impedance but are a little higher in capacitance.
Try checking out the Dunlavy's. I have NO personal experience with them nor am i familiar with all of their specs, but John Dunlavy has stated that they are the "most technically correct" cables on the market and offer a nominal 8 ohm impedance. Given that John is a true engineer in every sense of the word, they may at least be worth looking into. Sean
> |
If I'm not mistaken, InnerSound offers a 30 day trial period on their speaker cables, and you can return them if you don't like them. You might want to email them to be sure. Super nice folks to deal with ... Here's the email address: [email protected] |
I have had good results with alphacore MI2 biwire cables between an EAD powermaster 2000 and ML quest z's. |
InnerSound makes good high performance, reasonably priced speaker cables designed specifically for electrostatic speakers. I use them on my Eros system with great results. |
I own ML monoliths and have tried several cables. My favorite has been Audioquest Clear (no longer available new--but you can find it on the used market). In my opinion these were some of the best speaker cables made, although I have not tried everything and have only tried Purists interconnects--so I can not comment on their speaker cables. |
I own ML Requests and Krell electronics. I use Pure Note Epsilon Ref. Speaker cables, they have no cap. and low inductance. These were some of the few cables that sound real on my system. I tried Silversmith but they were too bright. |
I have tried many speaker cables on my Soundlab U-1. Best results were from Nordost Quatra-fil , Purist Dominus, Cardas Golden Cross and top end Audioquest, such as Volcano.
All of these are extremely high quality construction, low capacitance and use high purity metals in their construction. The ones I listed are the high end of each manufacturers line, so if cost is an issue consider finding used.
My personal choices (in order) is Purist Dominus, Nordost Quatra-fil, then Cardas. The Quatra-fil is the most transparent, the Cardas is the warmest and the Purist is the most phase accurate and prettiest midrange.
Observations made on Soundlabs may not transfer to your Martin Logan's. For instance, the midrange on your Aerius may make the Nordost a poor choice. I would not be surprised if the Cardas was the best choice with your amp and speakers.
Please post results if you try some of these and what you decide. |
Hello Dennis. Try the Kimber Bi-Focal"X" (9ga), or "XL" (5.5ga). This speaker wire is neutral, transparent with excellent focus and enveloping sound-staging (depth, width and height). This series posesses the rare quality of being neutral, quick and transparent with harmonic richness/realism and "body". A great match for electrostatic speakers.
peter jasz |
