"The alloys in the pipe would probably not have these elemental values but should be fairly close. I expect the steel conduit to be reasonably close to the iron value as well. So, a short run of lead to the soil is probably a lot lower in electrical resistance than the much longer run through the conduit."
The above is true but I think we need to look back to the classic definition of earth ground which is a copper rod 1,000 ft long inserted straight down into the earth. Theoretically this allows any am ...
One more point, an unknown - it's important to be sure just how well grounded are the water pumps along the way, from the reservoir/water factory to one's home. I have had an opportunity to see a friend discover his own apartment building water pump was probably not grounded at all. In view of the required power those electric motors need to have, this is no small problem.
Exactly the same thing happens when people use heating pipes for grounding; they discover more often than not that heating pipe installations are seldom really grounded. Thus, instead of receiving dead quiet, they end up receiving nice hum and eddy currents.
Theoretical ground, as outlined in IEC documents, should have a voltage potential of 0.8V or less and an impedance of 2 ohms or less. I have yet to see somebody lucky enough to have this, with perhaps the exception of those fortunate enough to own their own houses who have invested in at least 10 foot copper rods stuck into the ground, and without hitting or even touching any pipe down there.
This is the reasoning behind separate ground filtering, I think - you can't really have it, so you may as well be resigned to not having it and do what you can to make it as close to the ideal as you can. Simply isolating it is, in my view, sheer nonsense, because all too often, what you end up having is no proper grounding at all.
Hum loops are certainly possible, but invariably they appear when some ground point assumes the ideal ground potential and actually sees all but that theoretical value. In other words, a poor design job in which nobody bothered to look into separate ground potentials.
Here's a small practical example. Some years ago, I designed a very simple power amp. It worked, but the level of distortion was higher than I wanted it to be. Well, I worked by behind off trying to get it down, until a few months later it dawned on me that my feedback "ground" was at a much higher potential than it should ever have been due to poor grounding in my apartment building. A simple 10 ohm resistor between the NFB ground and the power supply ground ended that; however, I must add there is no math known to man which can work out which resistor will do the job, it's strictly heueristic, try and see. That 10 ohm resistor was about the 15th or 16th I tried.
Ground is the Al Quaida of audio, the insiduous and invisible enemy.
Cheers,
DVV
