how can a line cord affect frequency response ?

For RF to propogate it not only requires some type of conducting surface, but the physics of the medium (IE the physical environment that contains the RF) play an important role. Effective transmission lines do not power cords make. True certain frequencies of RF energy may attach to the surface of things other than metal, but in our case, audio amplifiers in metal cases, with power cords going in and cables I/O etc...which under normal conditions are shielded enough. I believe that very few home audio amplifers have ever fallen victim to any incident RF of enough energy as to be noticable.

This report of hard or glassy audible results from "RF" ...where does this come from? Have these types of claims been investigated? Have tests been run where an audio amp and cable have been exposed to swept RF sources while listening tests and/or controlled measurements of the device under test are performed? I highly doubt it.

Like the study of speaker cables, etc... it is a hard test to do and there is no shortage of opinions on the subject.

Remember, we are talking about engineered products. Yes we derive pleasure out of them but, there are cold hard facts based on decades of knowledge and years of experience put into the designs we love...
We need to stay true to the basic facts of engineering and device physics to truly have meaningful conversations about it.

Back to the question: How can a line cord affect frequency response?
In a typical audio amplifier power supply, on one side (we will call it the ac-line side) the transformer sees some resistance (from the dozens of feet of ac line going all the way back to the transformer on the pole), some inductance, and some capacitance.
-Aside: homes typically have 12/14 gauge un shielded solid copper wire that is the standard. Increasing the diameter of the power cord to say 10 or even 6 gauge does nothing to the instantaneous current available "out of the wall". SUre current densities may be different on fat and skinny conductors, but the end result will be the same. These lines are in series.

So the transformer, itself made of two massive inductors wound around some magnetic core decouples the amp-side windings from the line-side windings. This is good news. ANy dc offset on the line side, besides having ZERO effect from the power cord, can now NOT "get across" to the amp-side winding.

SO, on the the next stage: the recitifers. So now we have a sinusoidal 50/60 hz signal driving a bridge rectifer of some sort. You have AC going in and rectified AC going out of the rectifer. THis means that, if you looked at the signal out of the rectifer's positive terminal you would see the negative going waves are now inverted and you only have positive ac-bumps, spaced either 50/60 hz apart or 100/120 hz apart, depending on the overall design of the rectifier stage.
To "make" DC the rectifed pulses charge capacitors. With ZERO loading on the caps,(ie output transistors completely off) the charging currents from the rectifer gradually reduce as the caps reach full charge. The caps can only be charged at a rate directly related to the ac-line frequency.
In the typical class A or A/B output stage design...one of the three pins of the output transistors (either FET or BIPOLAR) are connected to this supply "rail" thus allowing useful operation of the device (I left out the other pins to simplify the discussion).

So now, imagine that the output transistors have a job to do and are now driving a speaker load. The current source for the output transistors comes from:
1) The resevoir caps directly -- during moments the recharging pulses from the rectifier are absent
2) The rectifer/power transformer directly -- During moments that the signal to be amplified places its demands on the output device during charging times. Under this condition the power supply current is not evenly split among recharging the cap and output device. Here is where a good stiff power supply carries the day for audio amps.

In the context of number 2-- consider higher frequencies that require amplification... it is very possible that higher frequencies cause related current spikes through the rectifer stages and back to the transformer/power cord/wall etc...
...also consider 'dynamic' signals such as drums requiring current during this window...

If you have a power cord with extensive ferrite beads or other ferrous material, it is possible that under these conditions, the higher frequencies may be diminished, due to increased resistance to the current at these frequencies b/c of the inductive response from the ferrite. This is why, I think, that ferrites in power cables may diminish dynamics or dull the high end.

It is more from a reducing-the-available-current-under certain-conditions effect than an actual designed filter result, though one can mimic the other.

Of course, the transformer itself has a frequency response and this may over-ride anything else by orders of magnitude under the high frequency analysis.

I hope this helps.

Regarding RF: I didn't mean to imply the line cord acted like an RF antenna when it was powered. What I was trying to say, however vague, is that it can act like an antenna and pick up RF energy. Although this energy should be eliminated in the transformer or perhaps some RF network in the PS input, there are cases where it can actually get into the amplifier itself and cause sonic problems. In this case, switching line cords may or may not make a difference. However, that being said, you still have lots of ROMAX in the walls which can pick up RF, too.

Dpac996: I believe you have bandwidth of the amplifier and bandwidth of the transformer, line cord, diodes, surge capacities, line droop, line drop, and dynamic compression all mixed up. Amplifiers simply don't work the way you describe. The power bandwidth of the transformer is typically no where near the amplifier bandwidth and sometimes it is not more than 47 toi 63 Hz. The line cord has wider BW to be sure, but keep in mind the only frequency on the line cord is 60 Hz, at least in the US.

Metro04: I do agree that materials do affect performance, but not like interconnects. Line cords only carry one frequency, 60 Hz here in the US. Interconnects can carry bandwidths from DC to 100 MHz with ease, if you use the right one.

Reb1208: Actually, if a line cord were to really seriously limit the power amp enough that it began to get warm and soft, it is more likely that dynamics will be compressed, rather than bandwidth. You might find that the low frequency peaks, not the high frequency peaks, is what got clipped. However, bandwidth below clipping, will be unaffected.

After writing that last night, it occured to me I should also elaborate a bit. When I said bandwidth, I was speaking in terms of the amplifier bandwidth, and I still state that line cords do not affect it. But since we are discussing line cords, let me relate the term "bandwidth" to line cords.

First of all, there is a voltage bandwidth and a current bandwidth but they are not the same. With respect to a line cord, the voltage bandwidth needs to be only a few Hertz wide, say 58 to 62 Hertz or so. Although the voltage bandwidth is much more than this, it doesn't have to be as long as the line cord can pass 60 Hz without loss.

The current bandwidth of the line cord, however, has to be much wider. In this case, a bandwidth from DC or a just a bit higher than DC to at least 600 Hz or so will do the job. The current BW can be higher than this without any problems.

I think I will start a new thread regarding line cord bandwidth as there is much to explain and it is really off topic from this one; probably next week sometime.

Spatial King,
I was describing the very basics of amplifer power supplies, paying particular attention to the rectifer output/resevoir cap.

I know that the power bandwidth of most amps easliy extends beyond 20 kHz.
Perhaps I did not articulate my thoughts precise enough, but I was not suggesting that an ampllifer's bandwidth is the same as the transformer.
I was purposely not getting into too much detail.
The question that Mr Tennis asked was how can ac-line cords affect frequency response. My simple answer is: I think that if there are ferrites or other inductive material involved in the construction of the cord, the inductance of this particular ac-cord may impede sudden fluctuations in current.

At the crest of the rectifed pulse, ie the charging pulse, the secondary of the transformer is now dumping current to the cap and the output device. It has to be this way or Kirchhoff's current law is full of hot air, which is not the case. The point I was trying to make is that during this brief moment the required current must ultimately come from the ac-line. If the ac-line is heavily inductive the instantaneous current performance may suffer.
I have never tested this theory on a real amp with different power cords. I have done circuit simulations in PSpice and i'm just throwing this out there as one way high frequencies or dynamics may be affected by PC's.

Spatialking,

What technical articals are you extracting your "voltage bandwidth" vs "current bandwidth" findings from, with regards to this topic and powersupply requirements?