Just to help the thought process a little bit. A true GROUND can’t have noise on it, it is zero volts @ earth potential by definition. What people refer to as ground noise, is a high DCR leg ABOVE ground and acts like a “signal” lead relative to true ground. There is a potential voltage.
We can make sure all grounds are the exact same zero DC potential (theoretically impossible as we have “wire” between the ground points of different DCR’s). The best solution is to isolate the signals ground plane to a small geographic spot where the wire DCR is mitigated. Some erroneously call this a single point ground, which is a larger global ground point.
Galvanic isolator can also work that break up a longer ground path and create a smaller local ground plane.
Another method is what I call passive RF absorber, make the impedance to a signal progressively higher as the frequency goes up. The expensive EPDM dielectrics used in ICONOCLAST acts exactly this way, the cable doesn’t pass RF well at all. This is clearly tested and shown in the data briefs.
The last method is brute force. Everyone asks what AWG cord should I use? For a super small current sink devices a smaller cord is fine. The voltage across the cord (resistance is the LENGTH of the cord) is proportional to the current. We can’t control the current but we can reduce the DCR so E=I*R is managed this way. Thus if we keep making the cord’s DCR lower won’t that make the residual voltage dropped across the cord smaller too? Yes, it will. One of the main topics of what is called BONDING is to make sure all the paths to ground (bond them together) are sufficiently low DCR. Using heavy power cords is a method of bonding the devices to the ground point. This also tries to mitigate ground loops to a lowest possible DCR such that they are inaudible.
Another method that is used is to force noise to go common mode and use a balanced circuit to remove it passively like when we use XLR cables. Some of these boxes do exactly that, they just run stuff through a balancing transformer or active circuit and make the signal balanced and the noise unbalanced and cancel the noise. Nothing wrong with that.
Before I bought a “box” I’d ask what method is used to remove the noise from the signal GOING to true ground so you understand the principals it is using. One or more? The problem isn’t the final destination at true ground, it is the PATH to that true ground.
A system with a true ground can’t have noise on it. Power cords with no safety ground (green wire) are dangerous as hell as the chassis could go HOT and when you touch it YOU are now HOT. Well, you’re hot for awhile until your dead body cools off. A device that has a three terminal plug is not double insulated (AC power is totally isolated from the low level “signal” circuit and chassis) and should never, never, never use a clipped ground power cable.
Most stuff the chassis is not double ISOLATED/INSULATED. A box in a box if you will. RCA usually use the metal chassis as the low level signal ground reference as an example. But the 120V AC isn’t using the chassis as the power GROUND (neutral white wire). The HOT AC (black wire) is on the opposite side of the white neutral wire power ground and not what we call earth (green wire). If the hot black wire or any HOT part of the circuit get across the chassis it is now HOT, not neutral. But With the green safety ground attached it will pop the breaker when it senses current on that wire, if you don’t remove it!
Being killed isn’t worth removing noise. Never, never pop the safety wire ground off your devices. Yes, at the circuit box the neutral and green appear the same, the thing to learn is that on the opposite end of the wire the PATH back to the circuit box isn’t the same. You never want the path back to be the green wire!
