There is a lot of necessary missing information in this thread.
A SPG, Single Point Ground is NOT a shield, it is an antenna and DOES put current in the "shield" that is now just a "wire" at this point. How do you think SPG works? CURRENT flows from the infinite open end to ground, thus current flows. A true shield NEVER, EVER has significant current flow. A SHIELD attenuates NOISE from inside to outside, or outside to inside (it works BOTH ways) as a RATIO, in dB, voltage. We should see virtually ZERO current.
FIX YOUR GROUND BUS!
If you use SPG, you won't die, but you are trading SIN for antenna coupled noise. BOTH are WRONG to accept as ideal as BOTH add CURRENT to the shield!
A lifted ground system with multiple lifted shields injects noise INTO the poor resistive differential (highest point to lowest point) ground system. As the ground is worse, the POTENTIAL difference from the end of the SPG ground point to the "ground" gets worse, aggravating the problem even more. Where is all that RF going? How is it getting there? What is it that current path parallel to? Maybe one SPG won't be an issue but many SPG added to a compromised ground will push it over the edge.
A SPG is trading off adding capacitive coupled RF noise injection into the core wires under the "shield" for removal of current induced SIN, Shield Induced Noise, from a poorly grounded system. A true shield has zip real current flow as the GROUND is at the same potential everywhere, or should be. By definition a ground at the exact same potential everywhere can't flow ground loop current.
A good shield will shield BOTH ingress RF and egress SIGNAL from a cable. That's what you want. Yes, our internal signal from digital devices needs to be SHIELDED from getting out of the cable as much as RF needs to be kept out. SPG "leak" RF into the exterior of the cable out as much as let it in.
One point to understand, is that the issues of a bad ground are superposition driven. As you "lift" a bad ground point and increase the resistance it looks more and more like an antenna, and less and less like SIN, Shield Induced Noise. They happen at the same time.
Once you REMOVE a ground point at either cable end for infinite resistance and thus no SIN current, we stop the SIN noise and fully replace it with Antenna coupled noise. A true SHIELD has ZERO current flow through it and is an ATTENUATOR. Making a SPG an antenna to stop SIN shield current isn't what best practice had in mind!
The ideal transfer impedance of the shield "R" times the shield current is the Voltage. Voltage equals current (should be near zero) times the transfer impedance at that frequency. But, when we have a bad ground our "R" goes far higher. Volts is current times resistance so that voltage product goes up as our shield tries to reach neutral state and flows current to the lowest potential spot in the system...where ever it is.
The ideal shield has the SAME resistance (zero) everywhere, same as a good ground and thus NO current potential to flow! It is all the exact same "spot" so current can't flow from HERE to THERE in the same spot. It can when here and there are DIFFERENT resistances relative to ground. This is why shields are rated based on their Transfer Impedance at specific frequencies. How LOW does the resistance look along the shield? Any "wire" will create a ground differential at frequency, so the right shield has to be used.
We can CHANGE the reference voltage from zero to ABOVE ground if we want to, but that EXACT voltage can't "move" or we again create a ground differential. We can have a one volt signal referenced to a million volts with a million and a ONE volt signal. This is still one volt potential difference UNTIL the signal reference point changes up or down. Usually we use something lose to EARTH ground for signal ground, not always. Connecting a SIGNAL ground to EARTH ground can mess this reference up. Make sure the circuit's "lowest potential" spot is understood. It may not be EARTH.
It is hard to remove a shield on an already built cable, true. So lifting a ground may be the better of two BAD situations. That's agreed to, yes. But we have to understand that it is WRONG from a SHIELD standpoint.
Many SPG systems with poor grounds worked better when the SPG were REMOVED and replaced with UNSHIELDED cable! This at least doesn't add noise to the ground bus and, XLR cables of good design can mitigate external noise passively with CMRR built-into the design. Try a UTP balanced cable and see what works. Many XLR have PIN 1 ground issue, too, so be aware of that. No, it isn't "earth" ground but "signal" ground. And yes, PIN 1 can cause SIN noise if they are too far apart in DCR.
NEVER, NEVER, EVER remove a safety ground!!! Now the fault current finds it path to ground through YOU! Touch a faulty unit with a "hot" chassis and guess what? Touch the next GROUNDED device and current goes through YOU. Sometime we learn a lesson ONCE and can repeat it to the angle at the pearly gates. This just tells you you need to FIX your ground, not make it a death trap instead.
Best, Galen Gareis
