Recently I've been following the posts on another forum of John Swenson, a design engineer for Sonore, and UpTone Audio (and perhaps others) on the topic of grounding, specific to devices powered by a SMPS.
In short, his research and testing shows that most of these SMPS units are inherently bad not just on a noise level, but also on a less understood and harder to test aspect known as AC leakage current. It was his supposition some time ago after an initial investigation, that AC leakage current from a SMPS is in many cases an equal or greater problem than just the plain switching noise itself. An example of this is the iFi iPower, a $49 wall wart with outstanding low noise specs (1 µV according to iFi), but rather large measured AC leakage current.
It's a somewhat in-depth topic and I won't attempt to fully summarize it here, except to say that previous knowledge and testing in this area (apparently dating back to at least the 1960s and therefore performed on LPS units as SMPS didn't exist yet) was always focused on low impedance pathways for AC leakage current. To this day, there is no off the shelf testing equipment that can easily measure high impedance AC leakage current, so John Swenson built his own custom tester including an ultra high impedance differential probe (around 10 Giga Ohms).
It turns out AC leakage current has not just a low impedance component, but can also exhibit very high impedance components on the order of 300 mega Ohms.
I'll stop there with my attempt at technical explanations, I'm not an EE, and some readers of this post might already be snoozing! Instead I'll mention the John Swenson authored solution which I've found to be transformative in using SMPS units in my 3 audio systems.
It's also worth mentioning at this point that avoiding the use of SMPS units in audio systems wherever possible or practical seems like sage advice, and thats not new or revelatory. An LPS will damn near always sound better, though it is larger, more expensive, and consumes more energy than these new fangled SMPS units do.
The solution John Swenson has prescribed is brilliantly simple and low cost, though it requires a small amount of DIY knowledge and skill. Lacking that, there is also a new off the shelf product that does the same thing as the DIY solution, albeit at $49 per SMPS treated.
John's custom test gear illustrates that ALL of this AC leakage current can be safely and easily shunted to ground with a simple connection to the DC output of the SMPS's negative leg going back to the AC safety ground.
I used a $6.50 Leviton AC plug, a $1.89 ring terminal, and some 14 gauge stranded copper wire at 50 cents per foot.
Finished cable looks like this:
Installed on the output of a Meanwell SMPS (at the input of the powered device):
This produced a rather shocking improvement in low level detail, dynamic contrast, stereo image size, even bass response, etc... all good, and no penalty.
This SMPS grounding technique has tamed detrimental effects caused by the one and only SMPS in my main audio system.