It absolutely does not matter is all your circuits come from the same leg or not. That is factually and electrically false.
While I don’t doubt that in some and perhaps many setups it won’t matter whether the components in a system are all powered from a single leg or from both legs, I would respectfully disagree with a contention that claiming it may matter in some or many systems "is factually and electrically false."
Following are excerpts from posts I had made in a couple of past threads which dealt with this question. (I’ve edited the excerpts slightly to remove wording that is not relevant here, as well as to remove links that no longer work):
Regarding the issue of splitting the load between two AC phases ... I looked through the ExactPower paper Jim (Jea48) referenced, the relevance of which is captured in its subtitle, "A practical guide for AV designers, installers, and electricians."
As an EE with extensive background designing analog and digital circuits (not for audio) I find the paper to be authoritative and credible. Which is to be expected, considering its authors. Among them, Henry Ott is a world renowned authority on numerous aspects of electrical and electronic design. Bill Whitlock (of Jensen Transformers) is certainly no slouch either. Some excerpts from their paper:Less than 300 microamps of ground loop current can cause hum as it flows in an unbalanced audio interconnect cable. However, harmonics of 60Hz that are generated from lighting dimmers or switch-mode power supplies sound like Buzzz mixed with a bit of Hummm and are more easily coupled by even smaller currents. Harmonics can add together when equipment is powered from different phases, so clearly there is an advantage to specifying same-phase electrical service to power the electronics systems in most cases....
Any leakage currents on the safety ground wires of split single phase load circuits fed by different phase legs will add together due to the 240V potential difference....
Power conditioners do not solve any of these common problems: Cross phase coupling (doubles hums & buzzes) .... What actually does solve them: Same phase power.Also, regarding ground loops ... it seems to me [Almarg] that if leakage current finding its way to the chassis (and safety ground) of a given component, via stray capacitance in the power transformer, EMI/RFI filters, etc., is out of phase with leakage current in another component that it is interconnected with, inter-chassis current flow between the two components, and therefore susceptibility to ground loop-related hum and noise, will have been maximized.
Again, I don’t doubt that in some and perhaps many setups it won’t matter whether the components in a system are all powered from a single leg or from both legs. But unless the system draws unusually large amounts of current and therefore might significantly unbalance the two legs if it is powered by a single leg (and keep in mind that most systems draw considerably less current than most hairdryers), it seems to me that what Henry Ott and Bill Whitlock have to say on the matter should be considered to be definitive as to what is best practice. And certainly not alleged to be "factually and electrically false."
Regards,
-- Al