I am finding differing opinions about a single point ground. Here are two discussions:
KY6LA Howard: In his paper at https://www.dropbox.com/s/kffp92esffo3zy5/How%20to%20Build%20a%20Quiet%20Station%20V2.pdf around page 21 - Use a single point ground
K9YC Jim Brown: In his presentation http://audiosystemsgroup.com/GroundingAndAudio.pdf around page 107 - Single point madness
Two respected sources with a difference in opinion.
My preference is a water pipe in the shack for a very low impedance connection between each station's "single point ground"; a Canadian compromise may be a "T" connection joining computer and rig with the long lead going to the ground bus.
One thing I don't see in most presentations is Kirchhoff's Current Law. If garbage escapes the box it must complete the circuit by some method even if that is a long path through many wires.
Stan - VA7NF
And If I've implied anything incorrectly, correct me before anything magical happens. I have no personal attachements with being wrong.
Ben Franklin just didn't know any better, but I bet he had suspicions. Better suspicious the superstitious.
I studied the Powerpoint and didn't find anything contradictory, but of course I couldn't hear the Presentation that went with it.
Jay - NO5J
The difference is the choice of topology.
Jim seems to be advocating a ring or Line ... While I am advocating a Star (or a modified bus)
Ultimately the intend is to keep the impedance and hence the potential noise voltages between the devices as low as possible.
In audio work, where Jim has much expertise, a ring works because the distances are shorter than the noise wavelengths
In RF, even short bonding distances are long compared to the wavelengths.
The ring topology and its variants tends to act as a huge antenna.. hence the Star or Modified Bus tends to have much better results.
Ultimately this is all theory.. and we can argue this until we are blue in the face...
In practice there are always local unknowns, couplings, paths, etc that can bite you in the butt....and change a star into a ring and a ring into a mesh and a mesh into a Star, etc.
Don't let Kirchhoff's Current Law confuse the situation. I am concerned with getting rid of RF noise... The current is conserved by converting it to something other than current at Noise Frequencies... Yes Garbage can escape the box but it does not need to come back at Noise Frequencies..it just needs to return as current....
> "Ultimately the intend is to keep the impedance and hence the potential noise voltages between the devices as low as possible."
Excellent advice. Despite our efforts to strive for one form of grounding topology or another, our intent is undermined by things for which we have little control. We almost always end up approximating the "mesh" structure that Howard shows.
Let's say we bring one large conductor from each piece of shack equipment directly to a ground rod driven into the floor of our shacks. That ground rod is further connected to a copper bulkhead outside the shack window where coaxial lines are bonded - and that bulkhead is tied to another ground rod that connects our grounded tower(s) in a ring that includes the AC power entry. We don't have, nor do we want a "single point ground." Ground should be established at many locations for the reason Howard states in the quote above.
Compounding the "single point" issue is that our equipment is further meshed with many grounded and interconnected RCA, BNC, and coax connectors. Oh, and what of the 3rd pin AC power cords that now tie the whole residence together in an even larger mesh? Almost forgot the grounded CATV coaxial bulkhead required by every cable TV operator (for those with CATV service). That coax forms another pathway into the meshed system.
When considering lightning avoidance, we definitely don't want a "single point" ground. We want low-Z grounding paths everywhere. We don't want lightning induced-surge currents to create a voltage divider in the grounding network. High Z ground paths result in surge currents that create a large voltage difference between equipment. What we want is a multiple-point ground where coax lines are bonded at the top of the tower, the bottom of the tower, and a low Z connection back to the shack cable entry and AC power entrance. Strikes on a tower need to divert to ground on a short path directly to the tower base. We want to try and a achieve a perfectly conducting, multipoint ground. This grounding method is used by every cellular telephone company.
For most conversations dealing with grounding, we would be well-off to avoid terms like "ground loop" and "single point ground." These terms have morphed in time and always come up because old habits, perceptions and theories are slow to die.
Ma Bell tried hard to teach me, but I got zapped plenty, and usually knew why. Survived anyway.
Jay - NO5J