I understand the reason some people prefer regeneration, others prefer filtration, others balanced transformers, etc is because the way in which the AC is filthy at different locations varies.And also because the sensitivity of different components to any given set of characteristics of AC noise and distortion will be different.
Unfortunately, IMO measurements of the AC noise and distortion characteristics are unlikely to be helpful, unless perhaps some unusually severe very specific issue is present. You may find the discussion between me and Mapman in this thread to be of interest. Some excerpts from my comments:
I suspect that in most cases even an oscilloscope would not provide useful information. While it would give a general idea of the overall magnitude of noise and distortion, it would say little or nothing about how the noise and distortion is distributed among what will inevitably be an enormous number of different frequencies. And there is no telling how a given component will react to a given noise or distortion spectrum. I doubt that even a sophisticated and expensive spectrum analyzer would be particularly helpful....And this from another member:
I think oscilloscope measurements would be unlikely to be useful because:
(a)The noise and distortion that is present on the AC will consist of a complex and probably time-varying mix of essentially ALL frequencies extending far up into the RF region, at many different amplitudes, including broadband noise as well as noise and distortion components at discrete frequencies..... There's no way to predict how a specific component in a specific system will react to that complex mix of differences.
(b)An oscilloscope won't provide much if any detail about that complex mix of differences anyway. It will pretty much just give an indication of the amplitude of the overall sum of everything, and perhaps also the amplitude of SOME of the discrete frequency components....
I learned many times over during my EE career that from a practical standpoint some things are inherently unpredictable, and can be determined only by trial and error. A good design will have as little sensitivity as possible to unpredictable variables, but no design is perfect....
Let's take the BrickWall surge suppressor/line filter that I use as an example. Its specs that are relevant to noise filtering, which are certainly measurable as well as probably being analytically predictable to a reasonable approximation, are as follows:
EMI/RFI Filter Response (bi-directional, wave tracking): With 50 ohm Rg load: 3db at 5kHz; 26dB at 100kHz; 38dB at 300kHz.
Let-Through Slew Rate: 5,000 volt/µs disturbance reduced to 28v/µs within AC power wave envelope, and less than 10v/µs outside the power wave envelope.
As with most designs, presumably those characteristics were chosen based on what the designers considered to be a preferable set of tradeoffs between the likelihood of audible benefit; the likelihood of undesirable side-effects; development cost; manufacturing cost; parts availability; physical characteristics; target selling price; marketability; profitability; utilization of available human, equipment, and intellectual property resources; etc., etc.
But can they, or any other manufacturer of such products, predict with any confidence what sonic benefits and/or side-effects will result when a product having those kinds of specifications is used in an arbitrarily chosen system powered by AC having arbitrary and unknown characteristics? The answers to those rhetorical questions seem clear.
03-20-14: Rodman99999Best regards, and best of luck.
To really view the noise, riding your AC lines; a Hewlett-
Packard 3577A Spectrum Analyzer is handy. Typical line
noise will be seen from 100Hz to 1mHz (and beyond).
-- Al