The deeper mysteries of 'ground'?

[Craig Sawyers]

The standard reference text on this is Grounding and Shielding, Circuits and Interference by Ralph Morrison, 5th Ed, 2007. I also have the 3rd Edition, which has a slightly different slant.

Anyway, he defines ground as a connection to earth, or its equivalent.

He then says of earth as a conductor "The earth is a complex conducting object. Resistances between two points on earth can vary from 10 ohms to megohms. The highest resistance areas might be on a block of granite, in the dry desert, or on a lava bed. The lowest resistance areas might be in damp soil or at the seashore. By burying copper conductors and treating the surrounding soil chemically resistances can be as low as one ohm up to a few hundred cycles. Typical wet earth contact resistance is of the order 10 ohms"

Also regulations do not allow two connections to ground - for example an earth spike connected to a radio chassis (and yes, I've done that; bad idea). A fault current that seeks earth can flow down the auxiliary spike and fail to trip the circuit breaker in the consumer unit.

[G8HQP Dave]

Maybe the aluminium horseshoes had a good layer of oxide?

[ms660]

I thought about that.

Regarding the electrocution would it be the fixings (penetration) rather than the shoes themselves I wonder, don't know a lot about race horse shoes but bog standard horse shoes are fixed with shoe nails.

Lawrence.

[Lucien Nunes]

Also regulations do not allow two connections to ground - for example an earth spike connected to a radio chassis (and yes, I've done that; bad idea). A fault current that seeks earth can flow down the auxiliary spike and fail to trip the circuit breaker in the consumer unit.

This is no longer generally true. Early earth-leakage circuit breakers, such as those historically installed by the supply company (DNO) before the consumer unit on rod-earthed (TT) installations, were often of the voltage-operated type (VOELCBs). These relied on current returning along the earth core of the installed wiring (CPC) to operate the tripping coil. The precaution quoted above was applicable to VOELCBs because additional earth connections would shunt the tripping coil and reduce sensitivity. Because of this severe limitation, they are considered obsolete and have not been installed for many decades. A few survive, often in non-functional condition.

Earth leakage protection in normal consumer units is provided by RCDs (residual current devices), formerly known as RCCBs (residual current circuit breakers) and RCBOs (similar but with overcurrent protection included). These all detect the residual current i.e. the difference between line and neutral current, without any connection to the earthing system. Thus they function correctly with any earthing configuration and will respond to earth faults, shocks and leakage via any path.

The main caveat with bringing in an additional earthing conductor of low resistance is that it is an extraneous conductive part that could, especially under fault conditions, be at a sufficiently different potential to all the interconnected metalwork in installation's equipotential zone, to cause a shock hazard. I'll return this and the difference between earthing and earth-bonding in another post, also circulating earth currents in a PME supply system.

Horses and other large four legged beasts are particularly affected by electric currents. Good heavy wet contact to the ground at a fair distance apart, even a few 10's of volts can be fatal.

[G4_Pete]

Re Cattle and horses.

My grandfather had the electricity removed his farm after a storm brought down some power lines and electrocuted some cattle, it was never re installed until after his death. I personally allways feel glad to see the horses alive and well after a violent lightening storm.

On a lighter note I know of two horses now that have figured out how to deal with high voltage on electric fences even though they are firmly grounded. They grip the insulated top of the posts with thier teeth, then pull up a few posts and hop over the wire, the grass on the other side allways being greener.

Potential gradients across earth often popped up in lightening and grounding training sessions.

Pete

[Radio Wrangler]

Where lightning has struck a tree or just the ground, there can be a circle of dead animals around it, caused by ground potential differences from the horizontal component of the current diffusion.

Some old cottages had been bulldozed and the land grassed over to make an entrance to a country park. Unfortunately the power cables hadn't been disconnected anywhere upstream and were in the earth just below the surface. Kids played there but nothing happened until a friend rode her horse in that way, something she'd done often, but the ground was a bit wet, the animal leaped straight upwards in a large convulsion and landed like a sack of bones. Quite dead. Instances of electrocution from ground (literally!) currents are fortunately very rare, and have rarely been witnessed. usually a lightning strike makes quite a distraction.

This is the reason for "Don't stand under trees" and "stand with feet together" advice in storms.

David

[G8HQP Dave]

I thought the advice was 'crouch down with feet together'?

[Billy T]

Originally Posted by mole42uk

I wouldn't believe too much of this story: " And because Fenix Two and Marching Song were wearing steel shoes, the current passed through their bodies unlike the other horses that were wearing aluminium plates. "

Any ferrous or non-ferrous metal: brass, aluminium, copper , silver, gold, and a myriad of other metals will conduct electricity (off-hand I don't know of any non-conductive metal, but I guess it is always possible) and high leakage to ground from a buried cable can not only give rise to electrical potentials over the distance between the hooves of a galloping horse, it can also electrocute a biped.

I work professionally with engineers etc when Hospital facilities and Universities are being upgraded or new buildings planned, providing advice to the Architects and Designers (including the Electrical Engineers) and recently,in a newly constructed multi-story building, I found neutral-earth links allowing currents to flow in copper pipes, plus a raft of other less serious issues such as cable trays acting as inductive loops.

On the positive side, the operating suites in a new childrens' hospital were perfect, though I nearly didn't get out alive as an Erk at the top of a ladder didn't make his presence known but lost his grip on a heavy steel plate and it landed at my feet!

In another matter, a new hospital was suffering serious corrosion in its heating and ventilation copper piping, turned out they were drawing salt water from a nearby estuary on incoming tides instead of fresh water from the tributary at low tide.

The point I wish to make is that professionals make mistakes and tradespersons misread plans, so errors in electrical cabling or damage (above or below ground) are not uncommon, so if making safety related checks, stick to the 'test-prove-test' principal, it may save your life.

Cheers

Billy

[Radio Wrangler]

A horseshoe, of whatever metal, fits onto the horn of the hoof, and the nails have a shaped point to bend their path outwards, so they stay within the wall of the horn until their curve brings them outwards. Horn is pretty similar material (and related to) our fingernails. It isn't a particularly good conductor. But, at the centre-rear of the underside of the hoof is a hairy, fleshy structure called the 'Frog'. The frog acts as an ancillary blood-pump, helping flow back up those long legs as the frog gets squeezed a bit by the force from the ground. When a horse is shod, the growth of the horn is cut back, so that with the shoe fitted, the frog will still be pressed by the ground.

Not even rubber horseshoes will give any protection from ground currents.

The bit about aluminium racing plates insulating is just daft.

There are, however, clip-on rubber/plastic boots. These were originally sold as a get-you-home spare tyre, but newer designs are sometimes being used as an alternative to traditional shoeing. These would act as insulation, but the whole thing would be wet and dirty anyway.

David

[Skywave]

Originally Posted by Craig Sawyers

Also regulations do not allow two connections to ground - for example an earth spike connected to a radio chassis (and yes, I've done that; bad idea).

Suppose the radio is connected to the house supply via an isolating transformer in which the earth lead from the house supply, although connected to the core and frame of the transformer, is not 'carried through' to the radio. The secondary of that transformer will then provide a balanced supply to the radio which can then have its own RF earth connection.
Yes, there will be some capacitance coupling, pri. to sec. in the transformer, but at 50 Hz, the resultant current will be very small, especially if the transformer has a pri.-sec. earthable screen.
I cannot see a serious potential electrical hazard in that arrangement, but I'll listen to advice from others here.

Al.

[Lucien Nunes]

Running any single piece of equipment from an isolating transformer is usually safe, if you want the chassis to float or to reference it to an alternative earth (e.g. one that works well at RF). However, it is an extreme measure for a radio which in most cases could be directly connected and earthed to the mains supply but given an auxiliary RF earth via a small capacitor, to limit 50Hz current through the RF earth. It may be undesirable to tie them solidly together for the following reasons:

#1: TN-C-S supplies.

The electrical supply to many newer buildings is configured as TN-C-S, meaning the neutral is earthed at source and shares a conductor with safety earth as far as the electricity meter, but the two are separate within the building wiring. It is almost the antithesis of star-earthing because the impedance of the protective earth conductor, which might be considered an equipotential with no voltage gradient nor any current (except leakage) flowing under normal conditions, is common to the entire load current. Instead of being a true equipotential, installations spread along the distribution cable (e.g. in different houses) will have main earthing terminals at arbitrary low voltages w.r.t. true earth, with a very low source impedance.

Extraneous conductors such as metallic gas and water pipes entering the building must be effectively bonded to the main earthing terminal (that's the heavy green/yellow cable by your gasmeter etc) to maintain the equipotential zone. These cables are heavy as the low source impedance can cause significant circulating currents despite the low voltage (remember it's caused only by voltage drop along the combined neutral-earth conductor leading to your installation).

If you create an effective auxiliary earth outside the resistance area of any existing earth electrodes (and it is actually quite hard to get below hundreds of ohms, but if you use a large grid on a damp day you might get dozens or lower), and you bring its potential into the equipotential zone where everything else is bonded or earthed to the main earthing terminal, then it can pass measurable current. If your auxiliary earth was truly excellent, say just a few ohms (highly unlikely!) you could get a few amps of neutral current passing into it as it will be in parallel with the neutral conductor back to the substation.

Many TN-C-S supplies use protective multiple earthing (PME) where the combined neutral earth conductor is effectively earthed in multiple locations. This helps keep 'true earth' and the CNE closer together throughout its length, by tying them more effectively in parallel. Nonetheless, if you have TN-C-S or PME supplies you should be aware that the likely small AC voltage from safety earth to true earth has very low impedance and can pump quite a current into your RF earthing system.

#2 coming next: fault conditions in the equipotential zone.

[Skywave]

Thank you, Lucien: very comprehensive reply.
In my particular situation, the incoming supply is overhead two-wire single phase with TT earthing, not TN-C-S, which - for what it's worth - I've always regarded as a rather dangerous way of arranging a mains distribution system. To wit, bonding the neutral to the 'earth'. But that's just my viewpoint, not a 'value judgement'.

Al.

[Radio Wrangler]

Thanks, Lucien.

Things get complicated somewhat when you obey the bit in the wiring regs about bonding all metalwork, water pipes and gas pipes. If you have metal pipes feeding you water or gas, then you have additional grounds. In the days of galvanised water pipes all screwed together, their impedance to the planet can be rather low, allowing for large fault currents. The adequacy of 6 square millimetre earth bonding wire seems doubtful.

Anyone who uses vertical antennae for amateur radio transmission is going to want to create a ground plane with a very low impedance. Now that is RF impedance and can be met with a raised groundplane/radial system with no actual connection to planet Earth. But if buried plates and radials are used, then the impedance at mains frequency can be quite low. For safety reasons, this needs either connecting to the main safety earth point with a hefty enough conductor, OR there needs to be carefully planned isolation.

David

[ms660]

So far as I'm aware an earth lead connected to an earth rod for RF reasons and brought into the building is classed as an extraneous metal part if any of it is exposed, even if a radio or whatever that earth is connected to is being powered via a mains isolating transformer and as such should be bonded back to the MET.

If I remember rightly the regs imply water and gas pipe bonding but not exclusively.

Just under 60 ohms ELI here with a DIY earth rod on a TT system, must be all the metal in the mining waste we're sat on

Lawrence.

[julie_m]

The point of TN-C-S is to build in redundancy, by having the neutral and earth bonded in as many places as possible. So there can never be much PD between them; and the loss of just one connection should not make a great difference, thanks to the low impedance already in parallel with it.

An earth stake positioned several tens of metres away from a house could have a potential difference between it and the house earth. This might not be enough to do you any harm, but could easily set up enough of a circulating current to melt any too-thin piece of wire between them. Hence why earth bonding wires are so thick: the current they may have to carry is not necessarily coming through your company head fuse.

It must also be bonded back to a point upstream of any RCD in the supply, with no other connection to Neutral downstream thereof, so every milliamp flowing out of the RCD can be accounted for as it flows back into the RCD.

My HF transmit antenna (10m vertical with ground plane/rods and an auto tuner) is half way down the garden. To be safe it is bonded with 10mm earth wire to the house MET (main earth terminal), that stuff is rather expensive BTW. Before connecting it I measured much less than 1 volt AC or DC twixt the two.

From what I understand 6mm earth wire is good for TN-S systems (which is what I seem to have looking at the incoming cable) and 10mm for TN-C-S so I went for the latter to be sure in case something has bee done to the cables under the road. Even my electric supplier couldn't tell me what I had!

[Synchrodyne]

Originally Posted by Skywave

Suppose the radio is connected to the house supply via an isolating transformer in which the earth lead from the house supply, although connected to the core and frame of the transformer, is not 'carried through' to the radio. The secondary of that transformer will then provide a balanced supply to the radio which can then have its own RF earth connection.
Yes, there will be some capacitance coupling, pri. to sec. in the transformer, but at 50 Hz, the resultant current will be very small, especially if the transformer has a pri.-sec. earthable screen.
I cannot see a serious potential electrical hazard in that arrangement, but I'll listen to advice from others here.

That arrangement might not “fly” in some jurisdictions. I had another look at the NZ regulations, and with an isolated supply arrangement, the appliance fed from the isolating transformer may not be earthed. For example there must be an appropriate level of insulation between earth and any earth socket/terminal in the isolated system. If this condition is not satisfied, then it won’t meet the requirements of an isolated system, and so would default to a normal system in which, if not double-insulated, the appliance would need to be connected to the mains protective earth in the usual manner.

As an alternative, it might be better to do the isolation in the aerial circuit, which could then have its own earth, although that separate earth should be bonded back to the main house earth stake – in fact such bonding is a requirement under the US regulations for example (see: http://www.reeve.com/Documents/Artic...ents_Reeve.pdf)

Isolation in the aerial circuit might also be better from a performance viewpoint. The receiver input is then effectively looking back into a physically closed circuit, and is probably less susceptible to picking up RFI from its immediate vicinity, including from the mains protective earth connection.

Cheers,

[G8UWM-MildMartin]

I avoid using the word in an electrical/electronic context, regarding it as an ill-defined foreign word that can mean chassis, mains earth, signal earth, analogue common, digital common, supply positive (mainly with PNP transistors), supply negative, HT- and a few other things.

[mole42uk]

Hear, hear!