Would this imply current delivery (from/to the cap) would be at best around the resonant frequency and at the same time amp stability (of the amplifier fed) could be compromised because of this?
Yes, in theory this can be true. In the real world, the fact that amplifiers function okay with power supply filter capacitors that have series resonances outside the power supply ripple and audio bandwidth tells us that the impact may be vanishingly small.
It also means that amplifier stability could be compromised in certain situations.
Regarding #1, wouldn't the frequency response of the stage be non-flat anyway, simply due to stray inductance and capacitance in the whole circuit, such that the contribution of the power supply filter(s) is comparatively negligible?
As I have stated in all of my posts, the impact may be vanishingly small, but the parasitic reactances may be additive in the overall scheme of the performance of the amplifier.
Yes, I understand that. What is not explained is what effect changing C1 parameters has on the output waveform, aside from modulating it with greater or lesser ripple -- or the shape of the ripple
waveform, in the case of high ESL -- superimposed on the DC supply.
We have already discussed this.
Again, in basic terms, it is about were the current comes from, where it goes and how it gets home.
Example.
Just for example to make it easy to see, let us say that at 15 KHz the impedance of the power supply capacitor has risen to 10 Ohms.
When the amplifier needs to produce X number of watts at 15 KHz it will not be able to. This will change the relationship of the higher frequencies to the lower frequencies. This will change the shape of the complex musical waveform at the output of the amplifier. This is just a basic example with numbers that make it easy to see what will happen.
This will not likely be the case with a well designed amplifier.
Remember that music is a complex waveform. Seeing a specific change in real time may not always be easy.
Again, in basic terms, it is about were the current comes from, where it goes and how it gets home.
This would fall under the category of power supply load step response in the frequency domain. I have already posted about this.
In real, real simple terms this is Ohm's law with an attitude (that is, it is frequency sensitive) and we call it impedance.
If we insert a 10 Ohm resistor in series with the output of the filter capacitor, the amplitude of all of the frequencies will be limited at the output of the amplifier. The waveform's amplitude will be less, but its overall shape will remain the same.
If we replace the resistor with an inductor, the amplitude of the higher frequencies will be limited, the lower frequencies not so much. This will change the shape of the complex musical waveform at the output of the amplifier.
The amount of change in my example would likely be easy to see using a basic oscilloscope. Changes at or just above the threshold of audibility are likely to be very difficult to see if we are looking at the waveform in real time. If we know the amplitude, frequency and phase relationship of every component of the musical waveform at a given specific time, we can us math to predict the overall shape of the waveform.
BTW, it you want to see a large scale example of this, just look at the output of your amplifier while it is playing music. Use your oscilloscope. Vary the treble control from minimum to maximum. This will show you how the waveform will change when the composition of the signal is varied in the frequency domain.
This should not be difficult to understand.
Again Dave, these are examples to show the mechanics in answer to your questions.