What length do you need?
18 responses Add your response
Acouple of months ago, a local HT dealer ask me to help him select the best speaker wire in the $5./foot range. He looking for a good sounding bulk cable to use in his HT installations. We try a niumber of different cables including the T-14 and Analysis-Plus. We found the IXOS 6006 Gamma Geometry Series cable was the best in this price
It is a great value in my opinion.
Buy some used from here or Ebay.
Unless you need a lot of wire, you won't need to spend $200 bucks.
It will sound different than your copper wire because of the silver coating, but is a very neutral wire overall.
The only comments I have ever read about deficiencies were slightly lean on low end.
I have it and don't agree with that.
I wouldn't go overboard on speaker wire. If you want to try out different things, spend some modest money experimenting with power cords, and IC's.
Marakanetz: The cable that you suggested is basically "generic" Monster Cable. Such a design is best suited for use below 300 - 500 Hz when connected to the typically inductive load of most dynamic loudspeakers. Above that point, you can expect roll-off to take place.
The only time i see cable like this being "suitable" for full range use is if someone were directly driving E-stat's with no step up transformer between the amp and the speakers. The high capacitance of the speaker would be somewhat "tamed" by the high inductance of the speaker cable, offering the amp what would theoretically be a little more of a "purely resistive" load. Since the nominal impedance of most "zip cord" type speaker cables is quite high, somewhere in the area of 100 ohms give or take, this would also help to buffer the otherwise low impedance that the amp would see if using lower inductance cables.
Cables are all about impedance matching. If you know the characteristics of the load that you'll be driving and the output impedance of the device feeding that circuit, doing a little math can take you a long way, both sonically and electrically. Sean
If I recall, you are using GMA Europa's? I exchanged e-mails with Roy Johnson, and he had two very specific recommendations as far as synergy with the Europa's. On the less expensive end, he suggested Analysis Plus. I took his advice and have been using Oval 12's with very good results. For a bit more, you could try the Oval 9's. And for an even higher end cable, Roy recommended Audio Magic Sorcerer's. The Sorcerer's are out of my price range, but I just picked up a set of Excalibur II interconnects. We'll see how they fit into the system.
I used T-14's for a long time, and I have to say that I prefer the Analysis Plus. Granted, my system has changed a bit since the T-14's were in, but I don't remember getting as much clarity across the entire spectrum with the T-14's. Not a knock at all, I thought they were great cables, but I would say the AP is a step up for not a lot more money. There are quite a few sets up for sale right now.
Best of luck,
Marakanetz: Dynamic speakers can offer various levels of capacitive or inductive reactance at various phase angles. E-stat's that are "direct drive" ( no transformer ) present a very high capacitive loads. I'm sure that there are other speakers that also offer less inductive loads.
As for the "frequency response" of various cables, Moncrieff went through and documented the results using various cables in IAR quite a long time ago. Due to the high levels of self-inductance, Monster cable aka "12 gauge zip cord" did not look pretty. The only cable that did worse was Fulton Gold, which is basically 4 gauge zip cord. It was linear up to about 125 Hz or so. After that, it rolled everything off due to the phenomenally high level of self-inductance. As such, it works GREAT as a subwoofer cable but sounds VERY "warm & dull" when used as a full range cable. It is for these reasons that so many of the "better" speaker cables strive for reduced inductance.
When it comes to speaker cables, the less inductance, the wider the bandwidth. This can be seen in the rise times of signals as measured at both the amp and at the speaker. High inductance cables will have a large overshoot on the leading edge of a high frequency square wave at the amp and will show a rounded corner at the speaker. Loading at both terminals becomes more linear as inductance is reduced but the amp may go into oscillation if capacitance becomes too high. As such, many manufacturers ( Nordost, XLO, etc... ) strive for a "reasonable" level of inductance with various levels of capacitance. Kimber's designs are less inductive and more capacitive, but not "crazy capacitive". Some other companies, primarily Goertz, Polk, Dunlavy, etc... threw caution to the wind and made very high capacitance cables. Their goals were to produce the widest bandwidth cables possible while delivering a very low nominal impedance. When using such designs, power transfer and loading characteristics are increased so long as the amp is stable. If the amp is not stable, one must take precautions i.e. use a zobel network either at the speaker or amplifier's output terminals. If one fails to do so with such a design, sonics will suffer and the amp may "smoke". Not a good thing, but some of my favorite amp / speaker cable combo's are of this very nature. Sean
Sean I realy love to see everything in numbers.
Let's take 20m of Alpha-core the high-capacitance wire or interconnect. Let's assume that such high-capacitance wire has 50pf/m(I think it's large enough) that will overall give a capacitance of 800pf and voice coil inductance of 0.3...0.7mH and determine a self-oscilation freequency of such load wich equals 1/sqare root of product of both.
Without precise calculation I'll state that in the denominator the number has a degree of -7.5 that states that self-oscilation may occur at tens of megahertz. These freequencies, if not filtered properly, may result a parasite oscilations of the amplifier that will act certainly negatively on normal amp performance. These freeququencies have also a large enough amplitudes so that it's lower freequency harmonics may also be amplified. The solution states in blocking such freequencies and limiting sencitivity for lower freequency harmonics to make the amplification or preamplification stable. If the amplifier was made tube or SS with carefull consideration of RF compatibility rules, wires are only to be considered as resistive load and should realy cost minimum. The best ones are only under zero-Kelvin temperature where super-conductivity effect was discovered.
You can also plug-in another reasonable values of capacitive and inductive reactance of speaker or speaker wire and all you realy need to know is simple radio equation where you equate both of the reactances to determin the self-oscilation freequency.
Every time you divide or multiply this derived freequency by 2 you get even-order harmonics and it means that you multiply the product of reactances to get a freequency divider. To get freequency multiplier you decrease reactances etc...etc...etc... Any harminics of self-oscilating freequency have smaller amplitude so by selecting a values of feedback resistor and capacitor you can always prevent amp's self-oscilations.
Please note that in tube amps output tube less-likely to oscillate by itself but small-signal ones realy do while in transistor amps both input and output transistors may oscillate at RF.
Oh, yeah! forgot to say that everything I say is towards nonsence of hassling with wires of any kind at audio freequencies. If the equipment poised to pick them up than no fancy snake-coated wires would help ever.
Old fassioned way sais "generic Monster Cable" is enough and sufficient and should only be considered as heavy duty, neccessary gauge and properly and tightly connected.
Marakanetz. Use whatever you like as speaker cable. As i've always said, you have to listen to your system, so you might as well enjoy it regardless of what anyone else has to say.
As far as your calculations go, your math and assumptions are way off. Other than that, there aren't too many amps that will actually make it out to a 1 MHz with any type of linearity ( equal amplitude output ), let alone a few dB's down. Most amps have hit a brick wall well before 200 KHz and are nose-diving at that point. As such, one would not have to worry about oscillation at all on "cheap" amps if what you said were true. Since i know that such is not the case in reality, and my ears tell me that there are differences in cables and one can measure the differene in how amps load up using test equipment, i'll stick to using what makes me happy, both technically and sonically.
If you can't hear a difference or are convinced that what you have is as good as it gets, more power to you. The fact that you are pumping all of your signal through an inductive output transformer that is of limited bandwidth really may make the differences in speaker cables far less noticeable / less important in such a system. If that is the case, one would be foolish to spend more money on something that offered very little return on the investment. Sean
I agree for my too optimistic values selected. they could go to 500kHz(yet forgot to factor by 6.28) especially in Alpha-core case managed to bring up unbeleavable huge p/u capacitance. The "generic Monster" or "generic RadioShack" has both capacitance and inductance much much lower than Kimber or AlphaCore.
Is it normal to oscilate at 200kHz? at what amplitude? If the design is "wide open" with no correction and with "zero feedback" than probably it will from even tens of microvolts.
There are circuit designs that are bandwidth limited ( on purpose ) so as to minimize the potential for high frequency oscillation. While doing this WILL reduce in-band performance to varying degrees, the manufacturer is more concerned about reliability and longevity than they are about sonics and being "technically correct". This is a perfect example of what we would call "real world trade-offs". Having said that, some designs are REALLY bandwidth limited and i'm not sure if this was done on purpose ( for added stability ) or if it is just a by-product of poor design.
As far as oscillation taking place at 200 KHz, i think that the Goertz Impedance Compensation ( Zobel ) Networks are tuned to come into play at 150 - 160 KHz ( give or take ). I can't recall the values of the parts used, but i know that they are different than what Nelson Pass has recommended in the past, what Jon Risch currently recommends and what Bob Carver feels is most appropriate. Obviously, what these highly educated EE's have to say should tell you something about how they feel about circuit design i.e. they all differ. I know that Carver's Sunfire amplifiers have impedance compensation built into them that comes into play at appr 80 KHz. Given the fact that the early Phase Linear amps used to blow up / sound horrible when connected to highly capacitive cables, i'm guessing that he's selected that frequency based on his past experience with trying to solve such problems.
As far as zip cord having a higher inductance / lower capacitance level, that is the very reason that the nominal impedance of these cables is so high at audio frequencies. As inductance is lowered and capacitance is raised, the nominal impedance is brought down closer to what the amplifier wants to see as a load. This has to do with what we call "transmission line theory" and equates to improved power transfer, more linear transient response and a wider bandwidth.
Most zip cord based designs will end up at somewhere between 80 - 120 ohms. This will vary with the gauge of the conductors and the amount of spacing between the positive and negative. This is very similar to most of the Nordost designs. Kinda funny how cables of similar impedances can give you such different sonics i.e. Nordost is lean and light in the bass / warmth region and "zip cord" is fuller sounding with rolled off highs.
When moving up to "fancy geometries", even Kimber, which is a relatively low inductance / higher capacitance cable design, clocks in at appr 20 ohms. This is for their heavier gauge cables with the Z ( impedance ) going up as gauge is reduced. The 4 TC/VS/PR series is closer to 40 ohms, give or take.
The only "audiophile approved" cables that i know of that are relatively current in production with a low nominal impedance would be Goertz and Dunlavy. Obviously, Dunlavy is no longer marketed under that name. For the record, Dunlavy started off as an RF engineer and then switched over to audio later in life, so he was well aware of transmission line theory and optmizing power transfer & bandwidth. While he thought that the differences in cables could not be heard during DBX conditons, he wanted to make a cable that he thought was "technically correct" for the specific job being done.
From a technical standpoint, both the Goertz and Dunlavy cables have impedances below 10 ohms, which is where you'll find the nominal impedance of most speakers at. The Goertz varies from 2 ohms to 10 ohms depending on the model ( heavier gauges are lower impedance ) and the Dunlavy's come in at about 6 ohms.
There have been other designs in the past, such as the infamous "amp blowing" Polk's, etc... that have also been quite low in impedance / high in capacitance. I'm sure that there are others currently being offered, i'm just not aware of them. I remember reading something about Coincident Technology and their cables being of a low nominal impedance / high capacitance design, but can't recall any specifics.
Obviously, some amp designs will oscillate at a lower frequency and have a lower threshold in terms of what it takes to excite them into oscillation than others. If one is in doubt, it is best to use a Zobel and not have to worry about it. Since some amplifiers already have Zobel's built into them from the factory, you might want to contact the manufacturer and see what they say. If a manufacturer is adamant against using such speaker cable designs, this tells me that they do not know enough about circuit design to help stabilize the circuit. It also tells me that the rest of the circuit is probably not very good either based on that lack of knowledge.
My experience is that using a low impedance speaker cable offers the best performance. Like anything else in the real world though, one may have to take precautionary measures to avoid the side-effects of such idealized designs when combined with wide bandwidth amplifiers of varying levels of stability. As i previously mentioned, some amplifier designs / systems may not benefit as much as others, so it would be foolish to to use a product that cost WAY more money yet offered no real audible or measurable results.
With that in mind, i would suggest that one should try some different cables out with various geometries / electrical characteristics and see what works best for them in their system. If you can't tell a difference, use the cheapest stuff that will work now and remain trouble-free in the future. Sean
PS... If looking at spec's for an amp and trying to ascertain the susceptability of the amp to various speaker cable / speaker loads, look at what is called "power bandwidth" and NOT "frequency response". The power bandwidth is the frequency range that the amplifier can deliver full rated power to while meeting distortion / linearity specs. Frequency response is typically rated at 1 watt, which tells you nothing about the amplifiers actual linearity at the frequency extremes. While it is true that narrow bandwidth designs ( limited power bandwidth ) can still oscillate, those with a wider power bandwidth are typically more susceptible to such problems.
As a side note, many manufacturers no longer disclose "Power Bandwidth" on amplifiers as it is a more telling specification in terms of how well the product is engineered. Instead, they concentrate on publishing the spec's that make their product look like a better performer rather than disclosing the big picture, which might expose "lower grade" designs for what they are. Sean
Sean, I like this good mentioning of "Power Bandwidth" is realy missing parameter that DOES specify how good this amp should sound to make everyone clear that specifications means something at least.
How about amps that have badwidth(and certainly amplitude) limitations truely according to a-must to follow RF compatibility rules and with no compromise to the hobbyist-built components? It realy just a matter of time fine-tuning input, driving and output stages to the perfection and carefully and thoroghly selecting values for correction to prevent any kind of self-oscillation. There are too many tips on amp-building and tuning but the bottom line is that self-oscillation on 200kHz or in the range is NOT acceptible on any component.
Certainly if OpAmp chips are used in input and driving stages the self-oscillation problem is already solved but sound is what will suffer. Tuning transistors or tubes is much more complicated and time consuming issue and time consumption is not good for business.