I'd like to interject some "technical" info and personal observations into this thread. As was mentioned previously, the logical train of thought is that a power cord should minimize voltage drop via minimal series resistance AND try to minimize noise. As usual, there are several different ways to do things with various trains of thought on the subject. Most all of them are valid, but some may be more effective than others given the specifics of the situation and design.
While the use of heavy conductors is common in aftermarket cords, one could also use several smaller conductors to achieve similar gauge size with greater flexibility. The use of several smaller wires also lends itself to configuring the wires in a specific array, i.e. twisting, braiding, spiral wrap, litz array, etc... These types of designs have several benefits.
First of all, all of the aforementioned designs are automatically less susceptible to RFI. In other words, it is less likely for these configurations to act as an "antenna" when directly compared to a common "side by side" or "stacked triangle" patterned cord. This is due to minimizing the inductance present in the cord itself.
Since these cords are less likely to act as "receiving antennas", the reverse is also true. Due to the various geometries and designs, these type of wire arrays also tend to reduce the amount of radiated magnetic fields and RFI. Besides the fact that our low level signals are being fed via interconnects that run in close proximity to our power cords, there is more to think about than just the magnetic field that is present. If you have a "noisy" device, such as digital gear, it is possible for the RF present in the device to be fed back down the power cord. In effect, the cord COULD become a transmitter of that RF signal to nearby cables / cords. Using "fancy geometries" can help minimize this situation and make it harder for the RF to "load" into the cord.
As was mentioned by other members, some cords also make use of ferrite cores for noise suppression. These work by TRYING to create an impedance mismatch. While we are normally trying to pay attention so that we don't run into impedance mismatches in most cases, that is not what works best here. By creating a wide gap in the impedance of the cable and the air around it, the ferrited cable effectively "chokes" the RFI out. Ferrite also works due to the fact that it is "modulated" by the common mode signal that is present from all of the wires going through it. Irregularities are "somewhat" filtered or absorbed by the transformer action of the ferrite.
Something to keep in mind though is that ferrite DOES saturate. Under heavy current conditions, ferrite can change value momentarily and / or permanently be altered. This is one of the reasons that ferrite is NOT commonly used on power amplifiers that pull large amounts of current. Due to the ferrite saturating, some keen listeners have noted a lack of dynamics, compression, smearing, loss of detail, etc... In other words, for ferrite to work best WITHOUT having side effects, it should be used in low current situations OR make use of a VERY large piece of ferrite to minimize the chances of saturation.
As to where aftermarket power cords are most effective, that is up for debate. I'll try to present a logical look at things and explain my thoughts so that you can follow along or know where i'm coming from in case you want to debate : ) Since we are using cords primarily for their lower resistance / greater current capacity / filtering action, let's look at what SHOULD be the most susceptible to reaping benefits from their use.
Since it is commonly accepted that digital gear is EXTREMELY noisy, that might be one place to start. The fact that the digital gear is typically a source component is also another factor. As was pointed out in another thread, "what is lost or polluted at the beginning can't be recovered at the end". With this in mind, my thoughts and experiences are that digital gear responds best to "fancy" power cords. This is NOT to say that other devices such as preamps and power amps do not respond to power cord changes, only that the results may not be as noticeable or consistent.
One of the reasons that changing the cords on digital sources is a two-fold benefit can be explained as such. First of all, we are dealing with a digital device. While the effects of minimizing noise going INTO any component should be obvious ( reason 1 ), try and reverse the situation. Since most digital sources have relatively small power supplies ( transformers, filters, filter capacitors, etc...), there is less isolation from their internals and the power line. As such, it is easier for noise generated WITHIN the component to "back up" into the electrical system. There are pulses of data being read and transmitted, clock frequencies being generated, etc... Much of this is also powered by some type of switching power supply. As such, the devices are NOT stable loads and introduce a lot of glitches, surges, spikes, hash and RFI into the electrical line.
This brings us to reason 2 why i think digital gear should be your first choice when replacing factory cords. By adding the increased "filter action" of a "fancy geometry" power cord, you can minimize the amount of information that is allowed BACK into the electrical system. In effect, this helps keep ALL of the other components from having to deal with the trash that otherwise might have been pumped into them via the common electrical connections. This results in OVER twice the effectiveness of changing cords on any other device, especially an analogue component. Of course, i'm assuming that the factory stock cord that was supplied with the unit is of at least reasonable construction and capable of supplying enough current without major voltage sag to start off with.
As to some further explanations as to why power cords may be less effective on power amps, preamps, etc..., the answer is quite simple. These devices TYPICALLY do not generate as much noise and normally have quite a bit more filtering in the power supply. While increased power supply reserve ( large filter capacitors ) and sturdier rectifiers ( turn AC voltage into DC voltage ) DO help the situation, RFI can still ride through all of this. In the radio communications field, it is common practice to superimpose DC onto an RF line or vice versa. As such, the increased filtering / heavy duty power supplies in these "non digital" devices does help to isolate it from incoming line noise and interference, it does NOT negate it or completely isolate it from some very specific elements. As such, the added "filtering" of a "fancy" line cord can only benefit the situation ( given proper current capacity ).
Some of the "oddball" devices that one might not think of as being "digital" or generating electrical noise might be amplifiers in the Sunfire product line and other "switching" type power supplies, some subwoofer amps ( commonly called "Class D", etc.. ) "digital" tuners, some newer "room correction devices", etc... As such, they are all likely to be candidates for power cord upgrades.
One last thing. Digital devices are typically VERY susceptible to "modulating" the power supply due to external excitation i.e. "vibrations". I have literally watched ( via a high frequency oscilloscope ) how tapping the case or shelf that a digital device was resting on generated noise back INTO the power line. Keep in mind that the designer of this device had already taken steps to minimize such things from happening. For example, the transformers were mounted on foam rubber supports, the chassis had damping compound applied in several different areas, ALL of the capacitors along with most of the other internal components had damping compound applied, etc... Even with COMPLETE attention to such detail, the ability to introduce noise via chassis excitation into the line was still evident. While most of this was more apparent going to ground, there was also noise being fed into the hot and neutral. Now you also know why many people suggest electrically isolating "digital" components from analogue components.
This also goes to show that many of the "damping and isolation" tricks that we do can amount to SO much more than what we initially thought or imagined. This also means that most of these devices will work best with their "lids" ON and lightly mass loaded. Besides helping to keep the internals and chassis from vibrating due to being modulated by airborne vibrations, the damped lid will also act as a shield. This will keep the RFI that the digital gear IS generating from being sprayed around the room and into other components.
Hope this helps answer some questions and gave you something interesting to read. As usual, i'm always open to comments or rebutalls. Please feel free to add to this or correct anything that you might feel is in error. Sean
>
PS.... Sorry for the novel, but i don't know when to stop : )
