From: ogicse!emory!kd4nc!ke4zv!gary@network.ucsd.edu Reply-To : gary@ke4zv.UUCP (Gary Coffman) Subject : Re: Proper J-Pole Grounding
>I assert that everything you do for lightening protection is a compromise. >You are trying to increase your odds that your house and equipment will not be >damaged by nearby lightening while trying not to bankrupt yourself in the >meantime. Additionally, I assert that shield braid stripped off old, useless >coax is a reasonable compromise.
Well I assert, from personal and broadcast experience spanning 30 years, that you can design a system that will handle *direct lightning strikes* on a routine basis. It takes some planning and careful layout, but it's not hard, nor is it overly expensive. At WXIA-TV, my other job, we take direct lightning strikes nearly every time there's a thunderstorm. Our downtime from such strikes is almost non-existant. The last time we went down from a strike, it was due to a strike on the power company's lines knocking *them* out, and our local generator picked up the load 12 seconds later.
I have a 440 MHz repeater located at the 970 foot level of that tower, and it has suffered no lightning damage in 4 years of operation, though there are burn marks on the repeater antenna from direct strikes. As in the shoemaker's case where his own children are barefoot, I took $10,000 worth of damage to my home station from a direct hit 3 years ago. That was because I didn't practice what I *knew* were effective lightning protective methods. After redoing the station following the hit, I've taken at least two more direct strikes, maybe more since I'm not home for every storm. There was *no* damage from the first, and the second got a Radio Shack UHF preamp mounted on my UHF TV antenna, but *not* the TV in the house thanks to proper protection. I could have protected the preamp too, but it wasn't worth the effort. A new one is cheaper than the protective measures required.
Since my disasterous strike, I've been campaigning vigorously to educate amateurs that you *can* avoid damage from direct strikes. The belief that there's no protection from direct strike damage is *myth*. It's equally myth that a blitz bug and a piece of #12 wire will protect you from nearby strikes. In fact, if you install supressors *slightly* wrong, they can be worse than no protective measures at all. Let me repeat that since it's vitally important, incorrectly installed suppressors can be worse than no suppressors at all. The instructions that come with suppressors, if any, are woefully incomplete and misleading. Ask an expert, or even better, become an expert on the subject.
The keys to effective lightning protection are surprisingly simple, and surprisingly less than obvious. Of course you *must* have a single point ground system that eliminates all ground loops. And you must present a low *impedance* path for the energy to go. That's most generally a low *inductance* path rather than just a low ohm DC path. Lightning energy is RF, and incredible voltages can build up across very small inductors when we're dealing with thousands of amps of surge current. The surge is very brief, so huge bussbars aren't needed, but the protective conductors must be low impedance at the frequencies of interest. Gary --
Date: Sat, 16 Oct 1993 14:42:02 GMT From: swrinde!emory!rsiatl!ke4zv!gary@network.ucsd.edu Reply-To : gary@ke4zv.UUCP (Gary Coffman) Subject : Re: Proper J-Pole Grounding
In article
Oh boy, this is a complicated subject. There are entire books published on it. One I recomend is _The "Grounds" for Lightning and EMP Protection_ available from Polyphaser.
At the risk of leading you into one of those "slight" mistakes, I'll give some general guidelines. First, a single point ground system is vital. Ideally you will use the entrance bulkhead as the single point ground. *All* wires coming from and going to your station, that includes phone, power, and control cables as well as antenna lead ins, must pass through this bulkhead, and all must be fed through proper suppressors that are firmly bonded to the bulkhead, mechanically and electrically. The bulkhead is then connected to your ground field by wide copper strap, 5 inch minimum, with no bends. The ground field can either be a broadcast grade radial field, or a sufficiently sized Ufer ground system. A single rod will likely do more harm than good because it's impedance will be too high to soak up the charge without excessive voltages. Letting your equipment "float" with lightning voltage is a viable protective strategy if you can guarantee adequate personnel isolation. This is a good way to deal with rocky mountaintop repeater installations where a good ground is unobtainable. The single point "ground" system will insure that there are no *differential* voltages across your equipment. So even if the equipment rises to several thousand volts above ground during a strike, no voltage differentials will exist that can drive damaging currents through the equipment.
Then there is the subject of cable routing on the tower, section bonding, plumber's delight antenna construction, etc, that needs to go along with all this to insure negligable ground loops in the cable runs. An insulated cable going up the side of your tower forms a half turn loop that will receive very high induced shield currents from the stroke current travelling down the tower body. You need to bond the coax shield to the tower in several places on the way down to short out this transformer action.
Blitz Bugs are practically worthless as protective devices. About all they're good for is to prevent the lightning current from burning down your house (maybe). They won't protect solid state gear at all. You need a properly engineered suppressor that mechanically bonds the shield to ground and that uses a sufficiently quick gas discharge tube to short the inner conductor when exposed to lightning voltages, but that will appear to be an open circuit to your normal RF voltages. There should be a DC blocking capacitor in the suppressor on the equipment side of the discharge tube. This will give the tube time to fire while the capacitor is charging, and thus keep the spike voltage out of your equipment. I recomend the Polyphaser NEMP line for this purpose. If you need DC pass capability, such as to power a mast mount preamp, a properly decoupled choke arrangement can be used to route the DC power around the blocking capacitor. The Polyphaser IN series is designed for this use.
I've only scratched the surface with this little discourse. You need to visit broadcast and two way sites and inspect how real systems are done. You need to read the books on the subject. You need to model your entire system and apply Kirchoff's Laws to it. Look especially at any inductances in the path to ground, and any mutual inductances between your mast and cabling, or between one cable and another. Figure on between 3,000 and 15,000 amps at a frequency of no less than 2 MHz for the main surge. Also consider that there will be reduced components all the way up into the UHF spectrum. Even very small inductances can develop large voltages under these conditions. Gary --
From: ogicse!emory!kd4nc!ke4zv!gary@network.ucsd.edu Subject: Proper J-Pole Grounding
In article <58440005@hpcc01.corp.hp.com> trapps@hpcc01.corp.hp.com (Stephen M Trapp) writes: >I am sorry is this is covered in a FAQ, but how do you achieve a single >point grounding system if the ham station is 40 feet from the fusebox? > >It seems the utility company wants a ground at the service entrance, my >ham station for RF reasons wants a short led to the ground near my rig, >the antenna wants the shortest, straightest shot to ground from its location. >I have linked the two ground rods with #6 wire underground to avoid safety >60 Hz problems, but short of a wide Copper bus around the house or ham shack >at the fuse box, there seems no perfect solution. By the way, my 8' rig >ground rod has a fan out of several 20' 14 ga. wires to help at RF.
Ok this is a common point of confusion so I'll try to address it here. The NEC mandates a ground connection at the service entrance. The phone company will also have a ground at their service entrance. Our tower base is buried in the ground, and may even have a radial system under it. These ground connections will all often be some distance from our amateur station. So this seems to conflict with the requirement for a single point ground, but not really. What we're trying to accomplish is to make sure that all the cables running in and out of our station are referenced to the same potential at a common entrance point to the station.
We do this by establishing what is called a *ground window*. Another term for this used in the commercial arena is the *entrance bulkhead*. *Every* wire that enters or leaves the station passes through this window, and *every* wire is connected to it, either directly in the case of nominally "grounded" conductors, or through an appropriate suppressor network in the case of "live" conductors. This means power, phone, coax, control cabling, *everything*. This window is then connected to Earth through the lowest inductance connection we can make. What we've done is to create a single reference point for our station, a single point "ground". As far as our station is concerned, any other grounds that may exist downstream of the single point connection have been effectively "shorted out" by the single point connection.
What we're trying to achieve is *equal potential* on all the cables leading to and from our equipment. Current can't flow between two points of equal potential. For equipment protection purposes, it isn't germane exactly *what* that common potential is. We don't care if it floats to thousands of volts above Earth potential during a strike as long as the *difference* in potential between our various pieces of equipment remains within tolerance. For personal safety reasons, and RF performance reasons, we want the Earth connection to be as good as possible, but we could actually not connect the ground window to Earth at all and still protect our equipment from damaging currents caused by lightning. Gary ------------------------------
Date: Wed, 16 Mar 1994 15:56:33 GMT From: ihnp4.ucsd.edu!library.ucla.edu!europa.eng.gtefsd.com!emory!wa4mei!ke4zv!gary@network.ucsd.edu Reply-To : gary@ke4zv.atl.ga.us (Gary Coffman) Subject : Re: Grounding and lightning protection
In article <2m4rsv$mba@bigfoot.wustl.edu> jlw3@cec3.wustl.edu (Jesse L Wei) writes: >Gary Coffman (gary@ke4zv.atl.ga.us) wrote: > >: Mobile Radio Technology had a series in the April and October 1988, >: and January 1989 issues about lightning prevention systems. They >: tried to give both sides equal time, but it's clear that the dissipation >: arrays are at best only minimally preventative, if at all. > >Now I have no technical expertise here, but I'd like to ask a couple >of questions: 1) Does the Corona effect prevent strikes, 2) do spline >balls work, 3) what about "feeler" charges? The reason I ask is that >Richardson Wireless Klub (K5RWK) had a meeting last night in which >a Richardson ham (I think) who works for TU Electric came and gave a >~1.5 hour lecture on lightning, prevention, and RF grounding. He >brought up some of the above-mentioned issues, and also said at the >end that he submitted an article to "one of the ham magazines."
First I want to note that I'm coming from the perspective of someone involved in protecting broadcast transmission systems, and as someone with lightning simulator experience. Also the local area has a thunderstorm frequency second only to Florida in the US. So I've seen a lot of strikes, and have a feel for what works and what doesn't. What I can't supply is much in the way of formal theory on the subject, only my reading of the trade press and a fair bit of other literature on the subject.
With that disclaimer out of the way, I'll give you my thoughts on your questions.
1) Corona, or point dischargers, are limited to about 20-60 microamps before streamer production begins. Streamers are the main mechanism by which near Earth lightning strikes are guided. So if corona breaks over into streamer production, you're going to attract lightning.
That's the principle on which lightning rods are founded. They generate streamers so that they are the preferred target of lightning bolts. Since they are installed with low impedance paths to ground, they are able to *divert* strike currents from harming other nearby structures. This is called the "cone of protection". It's diameter is equal to about 1/3 the HAAT of the lightning rod in most installations. (High towers have other problems, and a "rolling sphere" method of estimating the protective zone must be used.)
2) The idea behind "spline balls", and other dissipation systems, is to multiply the number of point dischargers so that currents can be shared so as to keep any one point's current below the streamer threshold. It's a good idea in theory, but in practice if the points are close together, their corona merges and forms streamers.
Remember that a typical strike is powered by about 20 coulombs of charge, and that individual points can't exceed about 60 ua without breaking into streamer production. So even if you have widely separated points to prevent merger, you still need an incredibly large number of them, especially if the cloud is capable of multiple strikes, which is the usual case. Also remember that cloud charge zones are in constant motion, and constantly inducing "mirror" charges in the ground below, so you don't have much *time* to discharge the currents safely.
The idea of a "protective space charge" is pure hokum IMHO. The winds in a storm are going to blow away any ions formed by corona as quickly as they can be produced.
3) I'm not familiar with the term "feeler charge" so I'll have to defer a response on that subject.
I'll add one more thought. There's a theory that if you can cause a *lot* of *little* lightning bolts, you can avoid the big dangerous ones. These "mini" bolts are supposedly so small that you can't see their strikes with the naked eye, but can measure them on a surge counter. This idea *may* work if the storm clouds aren't very energetic, and take *minutes* to build up to a strike, but I don't think it works in practice with the big thunderboomers we typically see. Gary ------------------------------
Date: Thu, 17 Mar 1994 06:29:11 GMT From: ihnp4.ucsd.edu!agate!howland.reston.ans.net!gatech!wa4mei!ke4zv!gary@network.ucsd.edu Reply-To : gary@ke4zv.atl.ga.us (Gary Coffman) Subject : Re: Grounding and lightning protection
In article <1994Mar16.162143.1@clstcs> armyrman@vms4.sci.csupomona.edu (Alex Myrman) writes: > >I too have antennas up on the roof and a couple long wire (dipoles) hanging >around off the house. >What should be done when lightning comes? I understand clearly that they >should NOT be in the radio but where should the lead-in's go?
Do commercial broadcast stations disconnect their antennas when a thunderstorm approaches? No. Do their antennas get struck by lightning? Yes, again and again and again. Do their transmitters sustain damage? Do their transmitter buildings burn down? Are their operators killed? No. No. And no. Why? Proper installation. (Truth be told, all of the above *have* happened at commercial broadcast stations, but in every case the cause can be traced to, you guessed it, improper installation.)
Proper installation isn't cheap or easy. Make the slightest mistake, cut the smallest corner, and you open yourself to catastrophic damage. So what's a ham with limited funds and knowledge to do? Many hams just disconnect their coaxes and drop them behind the radio. Some who are a bit more savvy stick the end of the cable in an old mayonaise jar. Neither trick is satisfactory. If your antenna is struck, there's going to be around a *million* volts on that cable, that much voltage can jump 100 inches in air, and it *will* if it has to in order to reach ground potential.
The only proper way to deal with lightning is to give it a controlled way to go to ground. It's going to go to ground one way or another, your only hope is to direct it in a way that's safe for you, your equipment, and your home.
>I have a heavy ground run to the radio room for grounding the equipment. >Should the antennas be connected to this, grounding the center conductor >and sheild? Should they be grounded and a real lightning rod be installed? >Or just disconnected from the radio's?
Well just disconnecting from the radio isn't good enough. You've got to give that lightning a *low impedance* way to reach ground. And that low impedance path has got to be able to successfully handle 4,000 amperes of *RF* current. That's what lightning is, nature's own spark transmitter.
Ideally you'll have a ground window installed at your station. (I know you folks are probably tired of seeing me preach about this, but it is the best protection you can have.) That ground window will have *every* wire that enters or leaves your station passing through it via proper lightning suppressors, including power, telephone, coax, *everything*. Note, arrange the cabling so that no down lead parallels an interior station cable run. Otherwise surges will be inductively coupled from the outside cable to the inside cable bypassing the ground window.
The ground window will be connected *directly* to your ground field by a straight low inductance conductor. In no case shall the conductor be less than number 8 solid copper wire, but should really be a wide copper strap, 5 inch copper flashing is good. (The reason wide copper strap is preferred is that it's inductive only at its edges, and because skin effect limits current penetration to only a few thousandths of an inch so you want as much surface area as possible.) Ideally there will be no bends in the ground run, but in no case shall there be any *sharp* bends. That adds inductance.
Note that in *addition* to the ground window, every antenna or support whose construction will allow it should have a separate ground conductor run to the station ground field. This will relieve the downleads, and suppressors, of part of the current load they'll have to carry during a strike.
A single 8 foot ground rod is *not* an effective ground field. Ideally we'd copper plate the Earth to form an effective ground field, but that's impractical. So we make do with driven ground rods. In average soil, a single 8 foot ground rod will have a resistance to Earth of about 230 ohms. That will place a connection to that rod at 920 kV during a 4000 ampere strike. Not good. As currents start to flow into the ground, the soil becomes temporarily *saturated* with charge. This limits the amount of current that can be quickly dumped into any individual Earth connection. So we need a bunch of Earth connections. How many is a bunch? Well good practice says that the total resistance to Earth should be less than 25 ohms, so that means at least 10 rods are required. How far apart should the rods be to avoid overlapping saturation zones? The rule of thumb is that ground rods should be no closer together than the *sum* of their lengths. That means that any two rods in the ground field need to be at least 16 feet apart.
The rods should be laid out in a star pattern with the rods connected to each other by no less than 1.5 inch bare copper strap buried not less than 18 inches below grade level. Note that these straps can be considered horizontal ground rods themselves and can reduce the number of driven rods needed in the system by about a third. So assume 7 rods, one central and six radial at a 16 foot separation. Make all connections to the central rod. That's your *single point ground*. Tie power company, phone company, and CATV grounds to this point as well as attaching your station ground and separate antenna grounds to this point. Never never never daisy chain grounds. All grounds must be tied to this single point, and only to this single point. (Note, if you have a tower, it can serve as the central rod. With its base planted in concrete, it forms a Ufer ground superior to a single driven rod. Note too that if you have metallic underground plumbing, that should also be tied to your single point ground by a strap connection.)
One more caveat. If your soil is dry sandy soil, or very rocky, you'll need more rods than for the typical case above. It's OK to extend your star out beyond the first ground rod, and in this case *only* it's OK to daisy chain along a radial from one rod to another, but more than two rods along a single radial reach a point of diminishing returns. The buried radials themselves, however, make a dandy groundplane for a vertical antenna and can extend out as far as you like.
I've left out many details in the above system, such as how to deal with bonding dissimilar metals, always making a *mechanical* connection as well as an electrical connection (solder *will* melt during a strike), what constitutes a *proper* lightning suppressor, etc. Entire books have been written on proper station installations. You should read at least one, _The National Electrical Code_. And I'd recommend one more, Roger Block's _The Grounds for Lightning and EMP Protection_.
Ok, that's the *proper* way to protect your station. Now what's the cheap ham way? Install an *outdoor* bulkhead panel near ground level and bring all your antenna coaxes through it with bulkhead feedthru connectors. Drive a rod into the ground at least 100 inches from the house and bolt a bar to it that has female coax chassis fittings attached, both shell and center connected to the bar. When a storm approaches, unscrew all cables from the bulkhead and screw them to the ground bar. This will keep dangerous currents and voltages *outside* your house. But that bar is going to reach 900 kV during a strike. Make sure there's nothing conductive nearby. Obviously *don't* ground the house bulkhead panel to this rod.
(Note that this cheap approach has several faults. First you've got to be home to connect the coaxes to the ground bar. Second there is such a thing as clear sky lightning. Not all strikes occur during a well defined storm. Third, any cable that passes parallel to the grounded coaxes is going to have a large surge inductively coupled into it. And fourth not all lightning is going to come into your house via your antennas. It can also come in on the power wiring, the phone wiring, or the CATV wiring. So this method should be considered a minimum *expedient* only. It does beat a mayonaise jar.) Gary -- Gary Coffman KE4ZV | You make it, | gatech!wa4mei!ke4zv!gary Destructive Testing Systems | we break it. | uunet!rsiatl!ke4zv!gary 534 Shannon Way | Guaranteed! | emory!kd4nc!ke4zv!gary Lawrenceville, GA 30244 | |