Driving a new ground spike

[Bo Deadly]

I haven't read this thread so apologies if this was already mentioned but there is only supposed to be one earth ground in a building. If you put a separate earth ground at the other end of the building for example, if there was a lightening strike nearby (inevitable), because the resistance of typical soil is higher than metal conductors, it very likely would travel up into one, through your stuff, down into the other and vaporize everything in between.

 

[nielsk]

There seems to be a popular misconception gathering steam lately, stated here as applying "grounding is about electrical safety, not 'noise'" to audio equipment.
Of course this applies to household appliances and machinery, but as anyone who has looked at a properly designed audio power supply knows, it's electrical ground is the 0 volt reference point in an audio device, and is instrumental in establishing the noise floor of a complex audio system, in addition to safety. While there are antique and modern devices with floating ground power supplies, suffice it to say that relying on the audio connection for Technical Ground is, at best, sub-standard and at worst dangerous.
The entire RF / EMI shielding system also relies on ground connections.
This is why there is a distinction between "Safety" ground systems and "Technical" ground systems.
Most audio devices when run by themselves do not suffer ground loops, but when incorporated into large systems with patch-bays and nearly unlimited interfacing possibilities, it requires proper Technical Ground systems and proper shielding termination, physical mounting and ground connections to ensure trouble free operation.
There is the popular analogy of an airplane "why does it not require a ground?"
Consider the fact that there is not a recording studio on the airplane, and it is not electrically tied to other businesses and homes on the same side of the service transformer.
Noise currents on the Technical Ground system that are allowed to flow into other devices rather than sunk to earth will appear in the audio signal.
The approach of multiple ground rods connected in series, with one end terminated at the Neutral bond and the other to Technical ground is an easy and effective way to minimize noise currents from flowing from the Neutral system into the Technical Ground system.
Accepted practice here is to determine the average ground water table and ensure the ground rods are driven to that depth. If this is impractically deep more exotic methods are employed.
While someone may plug in their laptop, interface and speakers into a outlet with no ground and not experience any ground related noise problems, it is wise to not reply on luck, but instead employ every aspect of a Technical Power system to prevent any related problems before they occur.
Who knows? They might get a pre-amp, and a compressor, and an EQ, and ....

 

[JohnRoberts]

I haven't read this thread so apologies if this was already mentioned but there is only supposed to be one earth ground in a building. If you put a separate earth ground at the other end of the building for example, if there was a lightening strike nearby (inevitable), because the resistance of typical soil is higher than metal conductors, it very likely would travel up into one, through your stuff, down into the other and vaporize everything in between.

That is an interesting hypothetical... Lightning is funny (not haha funny).

Up until recently most residences had a ground spike for the mains power junction box, and a metal pipe water main coming into the house. These days most such new construction plumbing is non-conductive PVC. I have never heard of lightning coming in through the plumbing and out the mains panel ground spike, but that doesn't mean it never happened. My telephone wire drop and even the satellite dish have their safety grounds bonded to the mains panel ground spike.

I am aware of two lightning strikes to my property, one killed a tall tree in my far back yard maybe ten years ago, that tree didn't fall by itself and I had to drop it. More recently lightning struck a tall pine tree maybe 50 yards from my house. My sump pump controller failed coincident with that lightning strike, but the replacement controller also failed without any lightning assist.

Speaking about lightning behaving funny, that downstroke in the pine tree, jumped over to a nearby utility line about 20' from the ground, presumably a lower impedance path to earth than continuing inside the tree. Curiously that pine tree does not appear dead, while the lightning caused obvious damage higher up the tree trunk.

Regarding earth's resistance or conductivity I can share an impromptu experiment. I was trying to use an old sump pump to pump water out of a pool of water created by my leaking water main pipe. I had to borrow an extension cord from a neighbor to reach all the way to the mudhole. I discovered after trying to use it, that "killer" extension cord I borrowed, had line (hot) swapped with the safety ground, so not only didn't the pump work, but the pump chassis was energized. I wasn't stupid enough to literally stand in the water while messing with mains power, but from standing next to the mud hole I felt a tingle in my feet from the energized pump. It wasn't enough current to trip a fuse/breaker, but enough for me to perceive, so my yard was not very low resistance.

JR

 

[dmp]

I haven't read this thread so apologies if this was already mentioned but there is only supposed to be one earth ground in a building. If you put a separate earth ground at the other end of the building for example, if there was a lightening strike nearby (inevitable), because the resistance of typical soil is higher than metal conductors, it very likely would travel up into one, through your stuff, down into the other and vaporize everything in between.

Having built a garage with a subpanel, this is wrong with my local code at least. You can put multiple ground spikes in and connect a single copper wire up to the panel. Any subpanels will also have a ground spike but at a subpanel the neutral will not be connected to ground. A subpanel could be at the other end of the building or in a detached garage (in my case). I don't think any of this is meant to deal with a lightning strike.

 

[abbey road d enfer]

There seems to be a popular misconception gathering steam lately, stated here as applying "grounding is about electrical safety, not 'noise'" to audio equipment.
Of course this applies to household appliances and machinery, but as anyone who has looked at a properly designed audio power supply knows, it's electrical ground is the 0 volt reference point in an audio device, and is instrumental in establishing the noise floor of a complex audio system, in addition to safety. While there are antique and modern devices with floating ground power supplies, suffice it to say that relying on the audio connection for Technical Ground is, at best, sub-standard and at worst dangerous.
The entire RF / EMI shielding system also relies on ground connections.

Talk about misconception...
As demonstrated by Grundy et al, technical ground continuity is not a requisite. Actually, connecting technical ground to Pin 1, which was common practice till the 80's, has been identified as a cause for EMI/RFI problems. Ensuring continuity of technical ground is not required at all. And not even shield continuity. There are many examples where disconnecting pin 1 at the receiving end actually improves performance.

This is why there is a distinction between "Safety" ground systems and "Technical" ground systems.

I don't think this is debatable.

Noise currents on the Technical Ground system that are allowed to flow into other devices rather than sunk to earth will appear in the audio signal.

How can "Noise currents on the Technical Ground" ..."flow into other devices"? That would indicate serious flaws in the design.

The approach of multiple ground rods connected in series, with one end terminated at the Neutral bond and the other to Technical ground is an easy and effective way to minimize noise currents from flowing from the Neutral system into the Technical Ground system.

I'd like to see how you can demonstrate that. Particularly how "noise currents " could flow "from the Neutral system into the Technical Ground system."

 

[Bo Deadly]

There seems to be a popular misconception gathering steam lately, stated here as applying "grounding is about electrical safety, not 'noise'" to audio equipment.
Of course this applies to household appliances and machinery, but as anyone who has looked at a properly designed audio power supply knows, it's electrical ground is the 0 volt reference point in an audio device, and is instrumental in establishing the noise floor of a complex audio system, in addition to safety. While there are antique and modern devices with floating ground power supplies, suffice it to say that relying on the audio connection for Technical Ground is, at best, sub-standard and at worst dangerous.
The entire RF / EMI shielding system also relies on ground connections.

Sorry but I'm going to have to push back on this nielsk. An earth ground is not required to achieve low noise. Most commercial pro audio devices sold today are not earth grounded at all because mains doesn't make it far enough into the enclosure to be concerned with safety. Consider that most USB audio interfaces are ungrounded plastic boxes and their noise floors are usually very respectable. The noise performance of an ungrounded device will depend almost entirely on the CMRR performance of it's inputs and outputs.

And RF / EMI shielding does not depend on ground connections as much as you might think. Electric fields are completely blocked by even a thin conducting surface like a foil as long as it's continuous and doesn't have any gaps or holes. With audio gear there are holes for connectors but the frequency of RF that is emitted into the enclosure will depend on how long the conductor is between the cable shield and the enclosure. This is why there are XLR connectors with a metal spike in the screw hole that bonds the cable shell and shield to the panel over a very short distance. Magnetic fields are attenuated by magnetic metals like iron and nickel more or less depending on the metal alloy, annealing, thickness and such. The enclosure does not have to be grounded for this to be effective. However, if it is not grounded the noise performance of the device will depend greatly on the CMRR performance of it's inputs and outputs which is not ideal. If you ground the enclosure, then the common mode noise is greatly attenuated relative to the 0V of the circuitry inside.

 

[Bo Deadly]

Having built a garage with a subpanel, this is wrong with my local code at least. You can put multiple ground spikes in and connect a single copper wire up to the panel. Any subpanels will also have a ground spike but at a subpanel the neutral will not be connected to ground. A subpanel could be at the other end of the building or in a detached garage (in my case). I don't think any of this is meant to deal with a lightning strike.

This doesn't seem right to me. Even if the ground spike was bonded to common piece of metal like a water pipe, that metal has inductance and so some current could be diverted through another path.

For example, let's say your jamming with friends out in the driveway and some people are connected to house power and some people are connected to garage power. If a signal cable somehow bridged the two grounds, a lightening strike could zap gear connected to the bridged cable.

 

[Speedskater]

Having built a garage with a subpanel, this is wrong with my local code at least. You can put multiple ground spikes in and connect a single copper wire up to the panel. Any subpanels will also have a ground spike but at a subpanel the neutral will not be connected to ground. A subpanel could be at the other end of the building or in a detached garage (in my case). I don't think any of this is meant to deal with a lightning strike.

Detached structures (like garages) often require additional ground rods.

 

[abbey road d enfer]

For example, let's say your jamming with friends out in the driveway and some people are connected to house power and some people are connected to garage power. If a signal cable somehow bridged the two grounds, a lightening strike could zap gear connected to the bridged cable.

In this scenario, where would lightning strike?

 

[Bo Deadly]

In this scenario, where would lightning strike?

Anywhere nearby could cause current to find its way to one ground spike but that is limited by resistance of soil so it might then also go through said "bridge" to the other spike.

See Grounding and Shielding by Morrison Sixth Edition section 5.23 on Lightning and Facilities which reads "If separate ground points are used, a nearby lightening strike can cause large ground potential differences between these utilities".

 

[Speedskater]

Detached structures (like garages) often require additional ground rods.

Just looked up the NEC reasoning:
Grounding and bonding is confusing. Since the Grounding electrode system (GES) {ground rods} is for lightning, we want a GES at the separate building as lightning wants to go to Earth. If the GES was only at the main building, that path is long and has a very high impedance. The lightning will find another path to Earth, causing damage and fires.

 

[JohnRoberts]

In this scenario, where would lightning strike?

Wherever it wants to...

Safety grounds can be redundant.

Anywhere nearby could cause current to find its way to one ground spike but that is limited by resistance of soil so it might then also go through said "bridge" to the other spike.

See Grounding and Shielding by Morrison Sixth Edition section 5.23 on Lightning and Facilities which reads "If separate ground points are used, a nearby lightening strike can cause large ground potential differences between these utilities".

In his second edition he shared that multiple ground spikes can steer ground currents to flow in wiring between the connected ground spikes as that wiring is likely lower resistance than soil resistance. He warned of magnetic fields (noise) from that current flow, not equipment damage. Redundant ground spikes for lightning management can used air gaps to provide some isolation while still delivering high voltage protection.

I don't know how common ground current flows are... apparently it is fairly routine for ground current flows around/between power poles and they are all ground spiked and ground bonded together.

Just looked up the NEC reasoning:
Grounding and bonding is confusing. Since the Grounding electrode system (GES) {ground rods} is for lightning, we want a GES at the separate building as lightning wants to go to Earth. If the GES was only at the main building, that path is long and has a very high impedance. The lightning will find another path to Earth, causing damage and fires.

yup...

JR

 

[abbey road d enfer]

Anywhere nearby could cause current to find its way to one ground spike but that is limited by resistance of soil so it might then also go through said "bridge" to the other spike.

I guess it depends very much on where lightning strikes.

 

[Bo Deadly]

Just looked up the NEC reasoning:
Grounding and bonding is confusing. Since the Grounding electrode system (GES) {ground rods} is for lightning, we want a GES at the separate building as lightning wants to go to Earth. If the GES was only at the main building, that path is long and has a very high impedance. The lightning will find another path to Earth, causing damage and fires.

That makes sense. I think there are multiple possible outcomes and like Abbey just said, it depends on where exactly the lightening strikes. If it hits the ground nearby, it will travel outward along the surface. So when it gets to a ground spike it will favor that entry to earth but if there is also an alternate path like a signal cable to anther structure that has it's own ground spike, it might follow that as well and zap stuff in between. If it hits the top of the structure, it may do the same but, because it has to handle the full force of the strike, it make also jump through the air to an alternate path and cause damage. So separate structures should have separate ground spikes. But each structure should have only one because all circuits within the structure are connected together. Although an exception to this would be if the structure has a metal skin like a post-frame building in which case the best option would be to have multiple connections from the skin to ground all around. Regardless, you should not have miscellaneous cables running between separate structures. Transformer isolation might limit noise but it wouldn't stop lightning.

 

[abbey road d enfer]

One thing to remember is that current does not flow only via the least resistant path; current is shared. There are thousands of Amperes there. 1 000A flowing in 0.1 ohm results in 100V

 

[JohnRoberts]

Lightning is funny... through air it appears to travel in roughly straight lines, but it is following paths of ionized air that are relatively conductive... after it reaches the ground it surely spreads out broadly but indeed tens of thousands of amps will leave a mark.

The power poles have a mature system of lightning management, back when I was still healthy enough to jog, I would stick near the side of the street with power poles, if caught in a lightning storm while jogging, hoping the poles would be protective.

JR

 

[shabtek]

We add ground rods to utility pole ground wire if soil resistance is high ideally more/deeper, but in rocky soils more rods paralleled. Some areas have poor grounding/bedrock and run delta, non multi-point grounded

In my application grounding is for equipment protection more than safety of us meat puppets.

 

[leadbreath]

If earthing is all about safety and not noise why do you need "clean earths" in high tech data centres where the computer romm is galvanically seperated with a1:1 isolation transformer and centre tap of secondary star winding is used as "clean earth"......?

 

[FIX]

I have built and installed hundreds of consoles, and ground or not, if the wiring and cables are connected correctly, you won't have noise. I had one install that had 90 volts AC of difference between the effects rack and console, when I got there, the studio had zero noise, it was one of my earlier API consoles, their solution was to leave a patch cable in between the lower bay and upper bay. Never a peep in the audio, as they installed it according to my plan. We found a piece of equipment that had a connection between the 120/240 switch that was touching the chassis. When we fixed it, the noise never got lower, but the death was eliminated...

We had a 1/2" thick by 6 X 6 copper plate with 1/4-20 studs for all the equipment in the room and studio.

My theory has always been, the console (or DAW) is the hub that everything has to connect to, so that is where the "house ground" goes to. I have never had good success with the obsessive "star ground" where everything goes to the ground rod. That method seems more like bargain rights that performance. Everything goes to the console, and it goes to the ground rod.

 

[leadbreath]

Ive measured over 100v between neutral + earth due to induced currents from harmonics caused by smps...