Gdnrbob, thank you kindly.
I believe I recognize which manufacturer is being referred to in the OP. Since the OP chose not to mention the name, I won’t either, although if I’m right it is certainly among the more respected manufacturers of audiophile-oriented power cords. I’ll say also that I would certainly consider it to be coincidental that its name begins with the same letter as the first name of the fellow who will soon be named NBA MVP for the second consecutive year :-)
That manufacturer recommends against particularly short lengths based largely on the rationale that more length rather than less will enhance the degree to which the designs of their particular power cords are claimed to be able to reduce high frequency noise that might otherwise couple from the component being powered into other components in the system, or even reflect back into the component which is producing it.
I suspect, though, that in most cases what Czarivey said is applicable, and a particularly short length won’t produce results that are any different than a more typical length of say 6 or 8 or 10 feet. However in cases where it **might** make a difference there are a number of tradeoffs that are involved, and how those tradeoffs will net out with respect to those length alternatives (i.e., whether a short length or one of those longer lengths would be preferable) figures to be dependent on the particular power cord, the particular component, the other components in the system, the voltage and noise levels of the AC at the particular location, and perhaps also on the room, listener, and even the recording.
For starters, noise reduction can trade off against bandwidth. More of one generally means less of the other, at least if everything else is equal or similar. And as Atmasphere has stated here in a number of past threads, power amplifiers in particular commonly require bandwidths on their incoming AC extending into the tens of kiloHertz, since they generally draw current in narrow spikes of very high amplitude, that occur during just a small fraction of each 60 Hz period. The manufacturer that I believe is being referred to, btw, emphasizes the extended bandwidth of their power cords, and even provides measurements to support that. So the noise reduction vs. bandwidth tradeoff might net out differently in the case of power cords from other manufacturers. And of course the importance of both of those factors is dependent on the design of the component being powered. And the importance of noise reduction that may be provided by the power cord, if any, will also be dependent on the designs of the other components in the system, into which noise generated by a specific component may be coupled.
Another tradeoff that is involved is voltage drop. The longer the power cord the more the voltage reaching the component will be reduced, although with a suitably heavy gauge cord that reduction will presumably be very small. But again, in the case of a power amplifier, especially class AB and class D designs, the wide fluctuations of their current draw that occur in response to the dynamics of the music, coupled with the fact that the draw occurs as spikes of very high amplitude within just a small fraction of each 60 Hz period, could make that a significant consideration in many cases. In fact in past threads Atmasphere has cited instances in which he has seen the voltage drop of a power cord reduce an amplifier’s power capability by amazingly large amounts. Although again there are application-specific dependencies, including the AC line voltage at the particular location, and whether that voltage happens to be higher or lower than whatever voltage the component was designed to sound best at (presumably 120V in most cases, for equipment sold in the USA and other 120V countries).
The bottom line, IMO: Within reasonable limits the length won’t make any difference in most cases. In cases where it does make a difference, that manufacturer’s length recommendation stands a greater chance of being applicable to their power cords than to other power cords, and the optimal length will also be dependent on many application-specific variables.
Regards,
-- Al
