Audiophile power cords vs. the cables in your walls

[marty]

Considering isolation does not affect it the same way, I could not say that. Why it happens is mysterious to me. The rest of the spectrum of playback does not benefit or sounds wrong and I can't recommend it for anything but lower range woofer/sub amplification.




Hey Marty,

The power from the wall runs at 60hz, so it always has 60hz on it. Other places in the world are on 50hz and I imagine it is similar.

All power cables create a field around them when they are being used. Certain shielding and wire configurations can change their shapes. Changing any of it will affect more than 60hz on the line (noise etc, technically speaking), but forms of negating other high frequencies has no bearing on this situation IME. There is relatively not much besides 60hz in the low range so I cannot suspect anything else is notable.

Ah, got it. I thought you were referring to conditioning with an external signal source. Thanks for the clarification.

 

[DaveC]

Just FWIW... The EMF surrounding the cable is related to the geometry of the wires, which also determines the electrical characteristics (LCR) or the cable. It's no secret you are looking for low inductance here, higher inductance results in a more intense magnetic field around the cable. Capacitance is less agreed on, but it is typical for capacitance to increase as inductance decreases. Resistance is determined by the cross sectional area and conductor material of course. Shields impinge upon the EMF of the cable and contain it, but this does have an effect on the signal and it's not good. This is mitigated by spacing and a cable geo with a compact EMF.

You can calculate by hand or sim LCR values of different geometries, build cables and hear the differences geo and shielding make.

For PCs and SCs, two cables that drive low impedance loads, thus have high current, geometry matters a little more than ICs. This is why I use ribbon wire. I can get very low inductance without excessive capacitance. If you experiment with braided geo you can achieve results that are clearly better than twisted pair or star quad, it's just much more of a PITA to build or manufacture. My geo has noise rejection like a star-quad with much lower capacitance. It's not better than the ribbon cable for SC and PC though, it is better for ICs.

Commercial cable has a lot of requirements related to abuse and durability, such as it being walked-on, driven over, etc. So most machine made cable conforms to those requirements and has only one conductor for each leg and the result is compromised electrical characteristics. Also, machine made cable often uses moulded plugs with very high resistance connections to the plugs, spot welds. Take a typical power cable and use it as an extension cord on your toaster oven and see how warm the plugs get. This resistance is included in the LCR and thus the effect the cable has on the signal.

The dielectric and jacket also make a difference, but that's not as relevant.

So for an audiophile grade power cable you really want to look into a hand made cable that has symmetrical hot and neutral with the ground external, as in wrapped around the hot/neutral cable, and uses plugs that use wire clamps. And if you use connectors with all the same materials you'll be way ahead of the game. Not gold.

Some would say all that matters is end to end resistance, but even here you want plugs with wire clamps and not moulded plugs, which will have variable resistance. But most component's power supplies are not as isolated from upstream effects as most assume. Boulder amplifier has claimed to make a power supply without cable dependence. I'm sure it's possible to make the effects inaudible, but that's not 99.999% of gear you'll be buying.

 

[lars66]

Boulder amplifier has claimed to make a power supply without cable dependence.

It is possible that the engineers in Boulder claim this because they are deeply convinced that it is true. If I were them I would try the Furutech Project V-1 Power, maybe they will confirm their opinion and maybe...

 

[Holmz]

It is possible that the engineers in Boulder claim this because they are deeply convinced that it is true.

Sure… and probable as well.

They are probably assuming that a power supply converts AC to DC and get the DC rails to have ripple that is low and not affected by the cable.

If I were them I would try the Furutech Project V-1 Power, maybe they will confirm their opinion and maybe...

I suspect that the engineers at Boulder, or in Boulder, (or ones under a potential rock), do not “have opinions”… but are operating in the realm of facts and objective analysis.

In a system… the boulder can be perfect, but I suspect that 2 different cords could be radiating different amounts of EMI to affect other parts of the system like ICs.

 

[lars66]

I fully respect Boulder’s engineering approach and the fact that their designs are based on rigorous measurements and objective analysis.

However, the assumption that a power supply fully decouples the equipment from the incoming AC line is only conditionally true. While DC rails may show very low ripple under static load, the interaction between the power cord, the mains environment, and the internal power supply is more complex under real-world conditions.

Different power cords can differ not only in resistance, but also in impedance vs. frequency, shielding effectiveness, grounding behavior, and their tendency to radiate or absorb EMI/RFI. These factors can influence how high-frequency noise enters or leaves the component, potentially affecting sensitive stages such as clocks, regulators, and low-level analog circuitry.

This does not contradict Boulder’s measurements; it simply acknowledges that some effects may lie outside the parameters typically captured by standard steady-state tests. From an engineering standpoint, the question is not whether the power supply works well — it clearly does — but whether the entire system, including cabling and environment, can be considered fully closed to external interference.

In practice, careful listening combined with controlled experimentation remains a valid complement to measurements, especially in high-resolution systems operating close to the noise floor.