- Written by Chris Groppi
- Published on 18 February 2008
On the Bench
I used my Smith and Larson Audio Woofer Tester 2 impedance analyzer to measure the impedance of both the speaker cables and the interconnects. This test is most applicable to speaker cables. The interconnects drive very high impedance loads with very small currents, and are therefore much less effected by cable impedance. I tested the interconnects anyway. I did not have the necessary equipment to measure frequency response, time domain response, or noise pickup, and was therefore not able to measure the filtering properties of the AC power cords.
For the tests I did do, I made some 10 Ohm loads for the speakers and interconnects. The speaker load is a five-way binding post with a 2% 10 Ohm resistor across the terminals. The interconnect load is an RCA jack with another 10 Ohm resistor across the terminals. The Smith and Larson woofer tester is designed to measure low impedances, so I could not use a high resistance load resistor to test the interconnect. In each case, I calibrated the impedance analyzer with the same loads used to measure the cable. This removes the impedance of both the load and the test leads. I then inserted the cable between the load and test leads, and did a 256 point impedance sweep from 20 Hz to 20 kHz. I used 48 kHz sampling, to try to avoid any roll off from the anti-aliasing filter. I did measurements with the cable laid flat, and also coiled up into about a 12" diameter coil. Measurements were performed on a 3m length of Silver Eclipse 52 speaker cable, and a RCA terminated Silver Eclipse 52 interconnect. As the geometry is identical, I did not separately test the Eclipse cable. I also tested my normal cable as a comparison, although I didn't include those plots in the review.
A perfect cable would show zero electrical resistance, and would add no phase angle at any frequency. Such a cable would have no effect at all on the load impedance, and I'd measure 10.000 Ohms resistance and 0 degrees phase angle at all frequencies. Since cables are not perfect, we will always see some additional resistance over 10 Ohms. In addition, any reactance in the cable will cause the phase angle to deviate from zero degrees.
Here (graph above), we see a plot of the Silver Eclipse speaker cable's phase angle as a function of frequency for both straight (solid) and coiled (dashed) geometries. The rise of the phase angle with frequency is a basic feature of almost any cable design driving such a low impedance load. The important thing to notice is the y-axis: the phase angle introduced by these cables is spectacularly low. Uncoiled, the maximum phase angle is 0.180, which rises to 0.220 when coiled. The small bump at 60 Hz is a measurement artifact from 60 Hz AC pickup. This result is a full order of magnitude (a factor of 10) less than my usual speaker cables measured the same way! In addition, the phase angle increases only a little when coiled. I did not plot the resistance, as it did not change much with frequency. The resistance added by these cables is also a spectacularly low 0.01 Ohms. This is also about a factor of 10 better than my normal cables. I really can't imagine better measurement results.
The interconnects show a similar story (graph above), but add even less phase angle to the load impedance. Maximum phase angle introduced is only 0.080. This is again about 10 times lower than my usual interconnect. The same small 60 Hz bump is there, again from the measurement setup. There was no significant difference between straight and coiled geometries, so I only plotted one line.
In addition to direct impedance measurements, I also measured the series inductance and shunt capacitance of the cables. I measured 300 nH series inductance and 3.3 nF of shunt capacitance at 15 kHz for the silver eclipse speaker cable. This works out to 30 nH/ft inductance and 330 pF/ft capacitance. The silver eclipse interconnect measured 135 nH series inductance (20 nH/ft) and 2.0 nF shunt capacitance (300 pF/ft). The characteristic impedance implied by these numbers is 9.5 ohms for the speaker cable and 8.2 ohms for the interconnect. This is fully consistent with the cable's very low measured phase angle driving a 10 ohm load.
Yes, these cables are expensive. But they made a real improvement in the sound of my system, even when compared to the very good cables they replaced. They also showed 10 times lower reactance, and 10 times lower resistance than my usual cables. This combination of great sound and great measured performance earns them my highest recommendation. If your system is already fairly polished, these cables (or some of Wireworld's more reasonably priced cables) could be the last little step towards audio nirvana.