Grounding for electronics

[karl riley]

Every now and then an article appears in one of our electrical journals which outlines ways to achieve very low impedance grounding electrode systems (GES). These articles are usually written by engineers employed by companies which sell the systems.

What reasons do they give for getting down to low single-digit ohms? One of the primary reasons given is the needs of sensitive electronic instruments. A response by one engineer (in a letter to the editor) to an article with this claim was that "The simple fact is that the GES has nothing to do with the performance of electronic systems." (He gives references to IEEE 1100-1992 standard).

Any opinions on this? We should really nail this down if we can.

Karl

 

[don_resqcapt19]

Re: Grounding for electronics

Karl,
I tend to agree with that statement. If all electronic equipment needed a connection to earth, we would not have airplanes, spacecraft, cell phones, laptops and other portable electronic equipment. Do interconnected electronic devices need a common reference? Yes, I believe that they do need a common point of reference in some cases, but see no reason that they need a connection to earth.
Don

 

[dereckbc]

Re: Grounding for electronics

Karl I have been in many warranty disputes with various equipment manufactures such as Lucent, Motorola, Nortel, Tellabs, Cisco, etc on this very subject. All the disputes revolved around lightning damage to equipment. The manufactures were not willing to replace damaged equipment, stating the GES did not meet 5 ohm?s or less as required per warranty agreements.

In all but two of the twenty or so disputes the manufactures lost the dispute. The two they did win out was because the two sites had separate GES, one for DC powered equipment and one for AC powered equipment. As you know they have to be bonded so we ate the losses.

All the other disputes they failed to demonstrate any practical reason for 5 ohm?s or less, once confronted with the high frequency impedance characteristics of single point ground systems. What was found was the SPG had been violated by their installation personnel by installing printers, video display units, modems, etc powered from outside the ground window. Other causes were AC conduit, floor anchors, etc., coming in contact with the isolated ground plane.

The only practical reason I can give is for RFI and other common mode noises. But extreme measure must be taken to get the point-to-point impedances down low enough to be of any use.

FYI: I am one of the contributing writers of ANSI/IEEE 1100-2004 aka Emerald Book

[ October 05, 2003, 09:39 PM: Message edited by: dereckbc ]

 

[bennie]

Re: Grounding for electronics

Karl: Do you have any specific excerpts from that standard?

 

[dereckbc]

Re: Grounding for electronics

Originally posted by bennie:
Karl: Do you have any specific excerpts from that standard?

I can give you one. "9.10.14 Recommended practice is to provide a resistance of 25 ohms or less for any made earth grounding electrode, but one does not need to seek unrealistically low values of resistance such as the popular 1 ohm value. Such connections are only used for electrical safety (not performance grounding) and not suitable use at HF. Recommended practice is to follow guidelines given in IEEE Std 142-1982[12] aka Green Book."

[ October 05, 2003, 09:52 PM: Message edited by: dereckbc ]

 

[hurk27]

Re: Grounding for electronics

Karl this would be a good subject for lightning protection. The one thing I see that is in error is the requirement for a low impedance connection to earth. Now we must ask what is a low impedance connection to lightning? We know that lightning has A/C properties and this would exhibit an influence on what the impedance should be. what is the most dominant frequency that lightning has when it strikes? We can also see this in the fact that lightning can or cannot see all available paths as some will seem as an open circuit to it even when it is of a very low resistance to the earth. I.E. when it strikes a 30' tree next to a 150' metal tower and leaves the tower alone. Of the studies that we carried out in central Florida (with the University of Gainesville) we found that placing several grounding electrodes of different types and different impedances caused the most strikes and provided the most effective path to earth to prevent damage. but a single electrode at one impedance would only provide a path if it matched the impedance the lighting was looking for. The more broad-banded the grounding electrode system the better the effectiveness of it.
It's design is not the result of a low resistance reading at an electrode, by far it incurs multiple calculations and the knowledge of Radio frequency in figuring the broad-bandedness of the electrode system to get the right answer.

 

[bennie]

Re: Grounding for electronics

I have finally found out the reason for a low impedance earth connection. At least this explanation works for me.

If Ben Franklin had stuck that key attached to the kite string, in the earth. There would not have been a buildup of electrons to charge a Leyden jar.

The ground electrode and related conductor will prevent a static charge from appearing on above ground equipment. The actual neutralizing of the charge will depend on the impedance of the bonding connection.

An inadvertent flash over on a PC board will destroy it.

An aircraft will build up a static charge. Should this charge be shorted and cause current to flow across sensitive equipment, there will also be considerable damage. Mini lightning strikes.

 

[pierre]

Re: Grounding for electronics

Hurk


My understanding of ligthning is it is a static discharge, one charge looking to discharge to an unlike charge (- to +).

That is the reason for cloud to cloud, cloud to trees, cloud to plane, etc...

Your example would only say to me the one rod was of the opposite charge the lightning was looking to discharge to.
That would be the reason the tree was struck as opposed to the tower, which we would think would be struck first.

Pierre

[ October 05, 2003, 11:21 PM: Message edited by: pierre ]

 

[hurk27]

Re: Grounding for electronics

Pierre:
Yes lightning is a DC static current when it makes it's connection with it's leader from the ground or cloud (depends on the polarity), But with one little side kick. It has A/C properties just like a set of points in an older car firing a coil. It has a fast rise time and will flywheel into a full sinusoidal wave with the frequency component intact. this also gives it the effect of being resonate to a certain impedance. Of course there wasn't to many high frequency scopes available at the time of our test but they are using time capture digital Oscilloscopes now and have proved this. but are still unsure whether or not the frequency composite of lightning is what steers the lighting to strike the tree instead of the tower.

If the charge in the tower was opposite of the charge in the tree why wouldn't there be lightning between these two points? the charge would of had to been enough to cancel out enough voltage to cause the lightning to go to the tree but with both planted in earth and the fact the tower had a grid ground planted under it I don't think there could of been enough of a voltage difference to cause this. as the earth would of shunted the two together.

[ October 06, 2003, 01:07 AM: Message edited by: hurk27 ]

 

[karl riley]

Re: Grounding for electronics

Great answers. This will all help me get it right in my book revision. Dereck gave the main IEEE reference from the article. The other was 4.6.1.

Hurk, do you have a reference where I could read about the AC characteristics of lightning? Your description of a variety of ground rods of different HF impedance reminds me of the different segments of a TV antenna.

My original question was actually not about the lightning function, though that was going to be the follow-up question. I was asking about the claim that a zero reference point in the earth is what the different electronics all need to reference to. I believe Dereck has told us that this is achieved not in a GES but in a grid within the facility. Do I have this right?

Karl

 

[dereckbc]

Re: Grounding for electronics

Karl I am tied up at the moment. I have made a few calls to other telephone operating companies, equipment manufactures, and some old friends in reference to the 5 ohm or less spec. Waiting for replies.

You got a handle on the grid question, I will expand later today.

 

[al hildenbrand]

Re: Grounding for electronics

Wayne,

The description of the bandwidth of the ground plane was a wonderful insight. Thanks.

 

[karl riley]

Re: Grounding for electronics

The mention of trees made me think, since they are frequently struck. Would the lightning have a choice of impedances depending on which branch it struck? Maybe the tree is giving a choice of impedances just like the varied ground rods. Just a thought.

Karl

 

[hurk27]

Re: Grounding for electronics

Sorry Karl as I have tied up and working late lately

But here is some of what the National Lightning Safety Institute has to say:

Grounding.

The grounding system must address low earth impedance as well as low resistance. A spectral study of lightning's typical impulse reveals both a high and a low frequency content. The high frequency is associated with an extremely fast rising "front" on the order of 10 microseconds to peak current. The lower frequency component resides in the long, high energy "tail" or follow-on current in the impulse. The grounding system appears to the lightning impulse as a transmission line where wave propagation theory applies.
A single point grounding system is achieved when all equipment within the structure(s) are connected to a master bus bar which in turn is bonded to the external grounding system at one point only. Earth loops and differential rise times must be avoided. The grounding system should be designed to reduce ac impedance and dc resistance. The shape and dimension of the earth termination system is more important a specific value of the earth electrode. The use of counterpoise or "crow's foot" radial techniques can lower impedance as they allow lightning energy to diverge as each buried conductor shares voltage gradients. Ground rings around structures are useful. They should be connected to the facility ground. Exothermic (welded) connectors are recommended in all circumstances.

Cathodic reactance should be considered during the site analysis phase. Man-made earth additives and backfills are useful in difficult soils circumstances: they should be considered on a case-by-case basis where lowering grounding impedances are difficult an/or expensive by traditional means. Regular physical inspections and testing should be a part of an established preventive maintenance program.


National Lightning Safety Institute


I believe much more study needs to be done.
Wayne

 

[dereckbc]

Re: Grounding for electronics

Originally posted by hurk27:

A single point grounding system is achieved when all equipment within the structure(s) are connected to a master bus bar which in turn is bonded to the external grounding system at one point only. Earth loops and differential rise times must be avoided. The grounding system should be designed to reduce ac impedance and dc resistance.

Wayne, I agree with the SPG theory. However they have left out some very important details such as the equipment has to be isolated from any incidental contact or unplanned grounds, and the power supplies must also be SPG through a "ground window".

There are very few companies who will implement a isolated SPG system, and those who do limit its use to only very specific areas were outages from lightning cannot be tolerated, such as digital switching equipment areas of a facility. A isolated SPG is very expensive to install and maintain, it takes special training for installation and maintenance personnel.

If you violate the isoalted ground plane with just a single unplanned connection like a missing floor isolation bushing or one RS-232 cable powered from outside the ground window, it puts personnel and equipment at great risk.

But I like the idea. I should, I have made a living out of it.

 

[hurk27]

Re: Grounding for electronics

Yes I have to agree with you on that Dereck. as any type of installation that requires the special knowledge of the equipment. needs to be installed or supervised by qualified personnel and not the average electrician. But the info that I was posting was some of what I had said before about the frequency component of lightning. the statement about the "shape and dimension of the earth termination system is more important than specific value of the earth electrode" is one that follows the belief that lightning cannot see all impedance grounds and seem to strike the most broad-banded one the most, And as Karl said (and might be onto something) as it would seem that trees are struck more often than power-lines that are just as close and of a lower resistance to earth. The reason I too believe too that the shape of an electrode is most important is because in radio it is known that the larger the diameter of the main radiating element in a transmitting antenna the more broad-banded it is. so if lightning exhibits the same properties then it would be presumed that this is why a grounding ring is effective or other large diameter electrode.

[ October 08, 2003, 12:01 AM: Message edited by: hurk27 ]

 

[hurk27]

Re: Grounding for electronics

The requirements in the NEC are mainly for the possibility of accidental contact from power-lines of higher voltages and to limit voltage potential between metal component's in and around a building. Lightning is just one of the things it also helps with but as we know it as designed it's not very good at it, as most can tell you that lightning can still do damage with even the best ground-rod it will still strike at the most dumbest place and sometimes not even seem to want to go to the earth. I had one house that the strike was to the roof then it jumped to a tree 20' away to me this didn't make any sense as there was even a bonded well casting and two ground rods at the service but thats what happened and it was witnessed by the homeowner pulling in his driveway.

 

[karl riley]

Re: Grounding for electronics

Hurk and Dereck, I am following closely. I am wondering if there is any source of accounts of lightning strikes that details the paths the strikes have taken, such as Hurk mentioned. Have there been any studies based on those forensic findings?

It seems to me that whoever writes standards tends to gloss over the unanswered questions.

Karl

 

[hurk27]

Re: Grounding for electronics

Karl I have been out of that area for a long time but I'm trying to do some research to see if there is any thing available on the net. I will get back with ya soon.

[ October 08, 2003, 10:50 PM: Message edited by: hurk27 ]

 

[hurk27]

Re: Grounding for electronics

Look at these Definition of Terms Used in Grounding By National Lightning Safety Institute.

This one in perticular:
"Resistance of Electrode:
Recommended IEEE practice is to provide a resistance of less than 25 ohms for any made earth ground electrode. Local conditions will vary this target figure. Figures of 10 ohms or less are standard practice in commercial codes and US Government tactical and long haul communications systems standards. Lower values, in the 1 to 5 ohm range, are useful only for electrical safety at dc and 50/60 Hertz . Lightning is an RF event, with typical HF characteristic impedances."


http://www.lightningsafety.com/nlsi_lhm/grounding_definitions.html

Also this one:

http://www.lightningsafety.com/nlsi_lhm/earth_electrodes.html

[ October 08, 2003, 10:55 PM: Message edited by: hurk27 ]