ground impedance, ground amplification

I had this idea a few years ago and every once in a while I come across a situation that makes me start thinking about it again. it happened again yesterday at a site I was consulting at.

this would be in the power system of a large studio.

there are 3 grounds available: A ground from the utility pole, the water main and a rod (or rods) driven into the earth.

there is a large 3 phase isolation transformer big enough to power all the recording equipment. the mechanical systems, HVAC, lighting, etc are fed directly from the line, upstream of the transformer, or from a seperate transformer. the secondary of the transformer has an isolated neutral that need not connect to the same ground as the mechanical systems. so you can ground this "technical earth" to dedicated rods, which presumeably have lower noise.

the problem is, the rods have a higher impedance than the water main and the utility pole ground. depends on the soil conditions. I have measured this in the past (with a borrowed ground impedance meter) and even with dozens of rods the water main is usually an order of magnitude lower Z. Does it matter? You can argue both sides. I am of the belief that you want as low a ground impedance as possible. I have actually witnessed a small difference in sound when switching grounds on such a system. with big soffit mounted speakers and high power amplifiers there is a subtle change in the low bass.

so what about artificially stiffening the technical earth? a large, high gain, high power AF amplifier could be used as an op amp. connected as a virtual earth amplifier: non-inverting input to the earth rod, inverting input is the new technical earth. small R of a few ohms from output to inverting input. there would need to be a safety system. something like a GFCI that when tripped would connect the earth rod directly to technical earth. you could possibly use the voltage across the feedback resistor to trip the safety system.

how big does the amp need to be? I *think* not too big, actually. this is to provide ultra low-Z at low currents (noise currents). probably impractical in big studio, but Im not sure. I think it would be easy to implement in a small system, say a mastering room. any thoughts?

mike p

Sounds reasonable, and with not a very big amp either. You'd have to calculate your signal out taking into consideration the voltage divider created by the resistance to ground or 'center tap' or whatever.

I've seen tube pre schemos where the K resistors are tied together and are floating from ground via a resistor. Adding an extra R to ground might make a voltage divider more stable, but at what cost?

Now, getting the exact signal that you want to remove from the power supply and phase inverting... wait - why not get a sine wave generator? Comes standard in any good UPS

I'm guessing that you if have dedicated circuits for your rack, your amps, and your computer/reel2reel then fine. Use the 60/0/60 for your preamps only.

A dedicated ground for your circuit should clean up that noise shouldn't it?

Then again, I'm not a power specialist.

[quote author="beatpoet"] wait - why not get a sine wave generator? Comes standard in any good UPS
[/quote]

you are onto me! For my mastering room I was hoping to be able to use a big 2 channel PA amplifier, one side doing the virtual ground thing I just described, the other side amplifying a sine generator to 60V RMS, feeding a 1:2 center tapped transformer to make 120V, the center tap tied to my new virtual earth. the output of this would be hardwired to the gear.

mike

Why not use the amp as a large class 'B' 60V amplifier and do 180deg phased 60V channels to a 1:1 with an isolated or floating ground for the secondary center tap. Do a passive transformer (impedance converter/phase inverter) & voltage divider network for your voltage noise NFB. That is, if you need one by that point.

I'm curious to see how if and how it works. I wouldn't plug your Neve console into it for a test run.

Are you sure it's not 120V peak you're looking for?