Anything beat dh labs t-14 for the money......

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
>

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.