1 square-inch copper to copper joint and have found it to handle a 25,500 ampere 8/20 waveform surge with no flash over and no change in resistance (0.001 ohms). ‘We have even wiggled the loose joint before and after the surge and experienced no change in its resistance. ‘The connection of a copper wire to a galvanized tower leg should be avoided even if joint compound is used. ‘The primary problem here is the low surface arca contact of the round wite with the (round) tower leg. Consider using two PolyPhaser TK series stainless stecl clamps, The TK clamps will help increase the surface area of the connection, as well as provide the necessary isolation between the dissimilar metals. Don't forget to use joint compound on exposed applications of the TK. clamps. For an even more effective connection, use copper strap in place of the wire with one TK series clamp. Silver oxide is the only known conductive oxide. (This is one reason why PolyPhaser's N-type coax connectors are all silver with gold center pins.) Copper oxide is not conductive and the proper application of joint com- pound will prevent oxidation. If copper clad ground rods are used, make sure the oxide layer is removed. PolyPhaser Corporation makes a Copper Cleaning Kit (CCK) that has all the necessary items (abrasive pad, joint compound, finger wipes and instructions) to make a good copper connection. Tinned wire should not be used together in the ground with copper ground rods. Tin, lead, zine and aluminum are all more anodic than copper. They will be sacrificial ‘and disappear into the soil. It is recommended that all ‘components be made of the same external material. Utilizing the above technique can make the difference between a site that stays on the air and one that requires ‘alot of maintenance after a short period of time. e 5 fee Z fig is MAGNESIUM 0.00 -O71 -161 1.93 -1.97 4 ALUMINUM 0.71 0.00 -0.90 1.22 -1.26 ZINC 161 090 0.00 -0.32 0.36 1 IRON 193 122 032 0.00 -004 E CADMIUM 197 1.26 0.36 0.04 0.00 S- NICKEL 212 141 051 0.19 0.15 nTIN 223° 152 063 030 0.27 g LEAD 224 183 064 031 0.28 © COPPER 2m 200 110 078 074 SILVER 317 246 1.56 1.24 1.20 | PALLADIUM 3.36265 «1,75 1.43 139 GoLo 387 316 226 194 1.90 | 1 Noble Metal Table: Accelerated corrosion can occur between unprotected joints if the algebraic diference in atomic potential is greater than + 0.3 volts ss é ° : = 2 Fo: ¢ § § @ 8 212-223-224 -271 317 -3.36 87 AAT 152-153-200 -246 -265 -3.16 051 -063 064 -110 -156 -175 -226 0.19 0.30 0.31 0.78 -1.24 -1.43 -1.94 0.15 -0.27 0.28 -0.74 -1.20 139 -1.90 0.00 -0.11 -012 -059 1.05 -1.24 1.75 0.11 0.00 -001 -047 -094 -112 -1.64 012 0.01 0.00 -0.46 -0.93 -1.11 -1.63 059 047 046 0.00 -046 -065 -1.16 4.05 0.94 093 046 0.00 -019 -0.70 1M 112 111 G8 218 60 -9I 175 164 «163 4116 0.70 051 0.00 18ae snot a BSESRROGE ence sires ere iti UFER PLACEMENT 2 Ufer ground is used, be sure the first ground rod is ‘= ast 2-feet from the tower base. Placing ground rods ‘some distance from each other gains the greatest benefit the money and effort spent. With widely separated sods, care must be taken to interconnect them properly. INTERCONNECTIONS As the spacing between vertical ground rods is in- ‘=reased, the interconnect wire will be able to launch or Sesk current into the surrounding conductive earth, ‘Therefore, it can be thought of as a horizontal ground “rod connecting to vertical ground rods, In highly ‘conductive soil, one should not be concemed about the ‘Sductance of such a straight wire because such a wire SBecomes like a leaky transmission line with very high Jesses. Therefore, wire size (skin effect) is of little S=portance, like that of rod diameter mentioned earlier, ‘2 long as it can handle the I - R of the surge. For Bighly conductive soil, a #10-gauge (bare) is the smallest wire that should ever be used. This type of soil is not common and in areas where soil conductivity ‘poor, such as sandy soil, the buried interconnecting ‘wire approximates the inductance of a wire suspended ‘air. This undesirable condition causes it to be highly ‘ductive. The high inductance prevents strike current (Gvhich has a fast rise time) from being conducted by the wire. Ground rods connected in such a fashion are useless se Ground Rod to 1-12" Copper Strap 88R-112S. 19 INTERCONNECTION MATERIALS To connect ground rods together in low conductivity soil, solid copper strap should be used. The strap may be as thin as 0,016 inch, For 1-1/2-inch wide strap, the cross sectional area will equate to a #6 AWG wire Greater thickness gains only a little advantage because high frequency currents (1 MHz) penetrate only to a depth of about 0.008 inch per surface, owing to "skin effect". The strap width should be about 1% of its length (e.g., 20-feet - 0.01 = 2.4-inches wide), providing a low inductance connection between rods. Connecting the strap to the rod may be done using a PolyPhaser clamp. For $/8-inch rods and 1-1/2-inch strap, the 58R-112S offers the best way to achieve a mechanically rigid, electrically superior, low-mainte- nance connection. The copper connection must be cleaned and a copper joint compound should be applied to prevent moisture. Clamps that bond straps together in the soil (MSC-3) and many other 1-1/2-inch strap to cable clamps (6AWG to 4/0) are available. 210 AWG 10 4/0 AWG. to 1-1/2" Copper Strap 200-1128. 4-172" or ° Multi Copper Strap sc-3GROUND ROD DEPTH The total depth each ground rod must be driven into the soil depends on local soil conductivity. Soil resistivities vary greatly depending on the content, quality and the distribution of both the water and natural salts in the soil. It is beneficial to reach the water table, but it is not necessary in all cases. In higher latitudes, rods should be long enough to penetrate below the maximum, frost depth. In most cases, a total depth of 40-fect or less is necessary, with the average being 15-feet. Depth also depends on the number of rods and the distances between them. KNOW WHAT YOU'RE DOING Be aware of the rod's progress when striking it in place, Rods have been known to hit a rock in their travels and ‘emerge out of the soil just a short distance away, In rocky soil, itis better to persevere in trying different lo- cations than to dig a hole and repack the soil around the vertical ground rod, ‘The repacked soil is less dense and will have a higher ground resistance. (Also, before the first rod is hammered in place, it should be measured as it is going in. This can tell a lot about the under layers of soil at the site, including when and if water or a highly conductive layer is reached. See Chapter 5 for ‘more information.) LIGHTNING, RF AND SAFETY GROUNDS AA single ground rod is never enough to ground a tower for lightning. Basically, there are three types of grounds, One is a ground for RF, such as an antenna counter- poise. A ground plane takes the place of the other half of a normal dipole. A good RF ground plane could be elevated above ground (tuned) and thus can not be a g00d lightning ground. If such a ground plane is prop- erly extended and placed in the soil, it will no longer be tuned. It can be an RF, noise and lightning. sink ‘Therefore, we can say that not all RF grounds are good lightning grounds, but a good lightning ground is a good RF ground for low frequencies. ‘The second type of ground is the lightning ground, This ground must be able to sink vast amounts of current quickly. The typical frequency range of lightning energy at the bottom of a tower is from de to the low VHF range (<100MH2z). The ground system must be a broadband absorptive counterpoise over this frequency range. ‘The third and final ground type is a power return or safety ground for ground faults. This is a low fre- quency (60Hz) ground and may be very inductive to lightning’s fast rise time, yet still be okay for grounding 60Hz. The electrical safety ground is often not a good lightning ground for that reason. A low inductance connection to a Ufer is, however, both a good lightning ground and a safety ground. Under a ground fault condition, more energy will be conducted to ground than during a lightning strike, due to the longer time required to clear the fault. Lightning has an cnormous peak energy, but the duration is very short. The Ufer has been proven to handle both without failure. GROUND THEORY Theoretically, one rod with a 1-inch diameter driven in homogeneous 1,000-ohm per meter (ohutvmeter) soil for ‘one meter would only yield 765 ohms. Driving it two meters into the soil would give 437 ohms. Going to three meters, however, does not give as great a change (309 ohms). One would get faster ohmic reduction and easier installation by using three rods, each one meter long, giving 230 ohms compared to that of one rod three meters long. This assumes they are spaced “greater than the sum of their lengths apart". If the interconnect- ing wire is also buried below the surface, then the ‘ground system may be less than 200 ohms. (Having ‘one deep ground rod, 40-feet or more, even if it reaches the water table, will not act as a good dynamic ground because the top 5 to 10-feet will conduct most of the carly current rise and will become somewhat saturated. This allows eddy currents to form in this top layer and ‘cause inductance to impede the flow of current to any further depth.)ming a in i | ' SATURATION ‘The statement that rods should have a separation, “greater than the sum of their lengths apart," originates om theory and the fact almost all ground rods will ‘saturate the soil to which they connect, A ground rod connects to localized, irregularly sized, three-dimen- ‘Sonal electrical clumps. Depending on the soil make- ‘= (layering, etc.) the cylinder volume of earth that a ‘ground rod can electrify can be visualized by this rule ‘ef thumb: The eylinder's radius and depth is equal to ‘Ge rod depth. This is known as the sphere of influence ofthe rod. Thus, the sum of the driven depths of two rods should be, theoretically, the closest that ground ‘rods can be placed. Anything closer will cause the soil {Glumps) connected in common to saturate even faster, Ground Rod Ground Level Incorrect Spacing @o) Correct Spacing Ft, Rod - BF, Radu Influence 101. Rod 10 Ft Radius Infuence ©)\©) 21 i i 5 s il ‘Theoretical resistance change for additionally spaced ‘ground rods EXAMPLE PROBLEM ‘Take for example a sandy area with a water table at the 10-foot level. The 10-foot ground rods are coupled and are to be driven to a total depth of 20-feet. A second rod is to be driven no closer than 20-feet to the first, but it should be 40-feet according to our rule! Our rule of thumb can be looked at two ways’ (1) Only 10-feet of each 20-foot rod is in conductive soil (the top 10-feet of each rod is in non-conductive dry sand), so 10-feet + 10-feet = 20-feet apart. (2) Without taking the water table depth into consid- cration, 20-feet + 20-feet = 40-feet.‘The answer: Following our rule of thumb, the #2 separation will not work! The interconnecting induc- tance will choke off the higher frequency components of the surge's rise time and create an L di/dt voltage drop. In other words, the surge will never reach the second ground rod. Following #'s spacing, the inductance will be less, but there are two other solutions to this real life problem, First, the interconnect inductance can be reduced by using copper strap. Second, by using chemical salts to increase soil conductivity around the rods and along the interconnect path, the inductance is also reduced. For the best performance, use both solutions together with #1's spacing. RADIALS ‘The interconnecting of ground rods is fine for those lo- cations where ground rods can be driven, Some loca- tions are rocky enough that only the interconnecting lines can be run. Horizontal radials, like those used on vertical broadcasting antennas, make an excellent RF and lightning ground system. Theoretically, four radials cach twenty meters (m) long, of #10 gauge wire, just buried will yield 30 ohms in 1,000-ohm/meter (ohm/m) soil. Eight radials would give about 25 ohms. Eight radials each fifty meters (163-feet each or 1,300-feet to- tal wire) on top of the ground or hardly buried could give about 13 ohms in IK-ohnvm soil... Theoretically, if we could drive rods, adding 2m long rods to this system, one on every radial (8 rods total), would calculate to below 10 ohms. If the rods were spaced every 10m on each radial (32. rods total), then the resistance would go to about 4 ohms. Remember this is theoretical and for 1,000-ohm/m soil, which is rather sandy or rocky. (We will see that a long radial run is not going to work as well as numerous shorter radials.) ROCKS AND MOUNTAINS. If the soil is rocky enough that the radials are sometimes in air while spanning between rocks, the accumulated inductance along the runs will choke off the surge currents, In this situation, numerous slightly shorter lengths, along with using solid flat strap radial conduc- tors, has been proven to work. The copper strap's sharp edge will concentrate the E fields that are present due to the existing L difdt voltage drop and breakdown or arc conto the surface of the rock or soil. There is, however a law of diminishing returns for radials. As with sprin- ler hoses, the amount of water, ot in this ease lightning nergy, at the end of a 75-foot length radial is such that ‘you may be wasting your time, effort and materials to {go much further. It is recommended radials only have a 75-foot run (no shorter than 50-feet if possible) and then additional radials from the tower be used to further reduce the surge impedance/ground resistance. This way the greatest amount of energy is carried off from the tower and away from the equipment hut, The radials runs should be oriented away from the hut as much as possible. (More on this later when we talk about fences and grounding other objects in the hut and ‘tower area.) The measured earth resistance of the radial system may be decreased. You will need to double ‘what you have to not quite halve the resistance value. On solid bare rock, straps will help spread out the charge on to the surface of the rock. Remember, 90% of the time a strike is an onslaught of electrons with like charge. They are repelled and want to spread out. In doing so, they lose energy due to the resistances involved, ‘Since little conductivity is present on dry bare rock, there will be little spreading in the time frame of a strike, If rain occurs before the event, then the rock will be quite conductive and the charge can spread out, loosing energy in the process. ‘The more it spreads, the more energy is lost as the charge density is reduced. ‘The Ufer ground works the same, it spreads the charge to the outer most edges. The use of the Ufer ground technique at the tower base and at the guy anchors will help. DOPING THE SOIL Salts may be added to increase the conductivity of the soil, but it is a temporary solution that must be renewed every year to maintain the elevated conductivity. Chemical ground rods (PolyPhaser’s PolyRod) can help capture the precipitation and direct it through the salts, creating a saline solution dispersed into the surrounding soil, The PolyRod can also be fed from a timed drip system, if domestic water is available. The ability toeof a FA ‘28d on copper tubing makes it easy to design an entire ‘Szund system using multiple PolyRods. The copper ‘Sbing electrically interconnects the PolyRods and acts, ‘25 2 means to disperse the saline solution along its ‘Seth. The surrounding soil is then made conductive, ‘seducing the inductance of the copper tubing. (More on ‘chemical doping in Chapter 13.) (OTHER MATERIALS ‘Grounding materials have appeared on the market ‘elaiming great things. They indicate that after applying, “Se resistance never changes. They also imply the earth ‘sistance is invariant. No so! It is the material's ‘sistance you put in place that will be the same over ‘Sie, not the real earth resistance! 2Bchapter 4 Guy Anchor Grounding Guyed towers are better at handling lightning surge currents than self-supporting towers. This is only true if the anchors are grounded properly. Some of the strike current traverses the guy wires (instead of the tower) and may be safely conducted into the guy anchor ground(s). With some of the current conducted by the ‘guy wires, less is available to saturate the ground at the tower base. TURNBUCKLES Tumbuckles should never be relied on as a path for lightning current, If the turnbuckles are provided with a safety loop of guy cable (as they should be), the loops may be damaged due to arcing where they come into ‘contact with the guy wires and tumbuckles. The following diagram shows the preferred method of grounding the guy wires - tying them together above the loops and tumbuckles. m4 U-Bolt Clamps 1/4" Galvanized ‘guy wire Optional aux. power supply — with Zine block Galvanized Stoo! a Ss coreow sa! Zine block with aux. power supply use magnesium block with no power supply Side view of a properly grounded guy anchor. Cor- rectly dressed and installed, the ground wire prevents ‘surge currents from welding turbuckles and damaging the safety loop.——s “asse connections should not be made with copper see. The reason - when it rains, natural rain water has = pH of 5.5 to 6.0 which is acidic. Copper is only at- ‘sacked by acids. Dripping water from the top copper ‘sere will carry ions that react with the lower galvanized ==>) guy wires. The reaction washes off the zinc ‘seating, allowing rust to destroy the steel guy “The best way to make the connection is with all galva- ‘seed materials. This includes the grounding wire or ‘cable and clamps, ‘Eest above the earth's surface, the galvanized wire is ‘bended to a copper conductor that penetrates below ‘ace to a radial system which spreads out the strike “ergy into the ground, “How high this bonded connection should be placed ‘above the soil, depends upon local snowfall or flood ‘eves. Snow's electrical conductivity, although low, “<= still cause battery action from the copper through “Se surface water to the zinc by the solar heating of the ‘sires. The joint should be positioned above the usual “Seow or flood level 25 The lead is dressed straight down from the highest to the lowest guy or with a slight tilt toward the tower at the top. After bonding to a guy wire, it should be dressed downward from the lower side to the next guy wire, The most effective way to bond is to lightly wire ‘brush the members, removing all oxides, and then apply a joint compound if a pressure clamp is to be used. To ensure no arcing will occur through the tumbuckle, a connection from the anchor plate to the ground rod is recommended. Interconnect leads that are suspended in the air, must be dressed so the bending radius is not sharper than eight inches. For guy anchors in typical soil conditions, use similar tower base radials, inter-spaced along their length with ground rods. The radials need not be much longer than 20-feet each because of the lower currents due to the high guy inductances. Two radials are the norm. If a wwire/chain link fence is used, radial should bond to the fence post with a PolyPhaser TK style clamp and con- tinue its run to 20-feet.Se Ground Impedance chapter 5 —_ GROUND MEASUREMENTS For towers, with or without guy anchors, we need to de- termine the type, number and length of the radials required for a good lightning ground system, To answer ‘these questions, the soil conductivity must be known, MEASURING YOUR SOIL A simple method for determining soil resistivity is shown below. Four equally spaced electrodes are driven to a shallow depth; the penetration depth (b) is kept small in comparison to the inter-eleetrode spacing (a) where a > 20(b). A known current is circulated between the two outer-clectrodes while the potential is ‘measured across the inner pair. If the electrode spacing (a) is in meters, use the formula: & p=6.28a I 26 fess] mayeee cy i Pe ‘ Ce Tbr trier fe "Fy ern | Ley ‘+208 (APPROX) Four stake method of measuring soil resistivity, This gives the soil resistivity in ohm-m, Generally, the electrode spacing (a) corresponds to the depth of soil seen by the test current. By varying the clectrode spacing, a profile of resistivity versus depth can be obtained. Another way to obtain a profile of the soil is {o measure a ground rod as you hammer it into the soil Plot the measurements to see if any major changes occurred over what is expected from simple calculations.‘MEASURING YOUR GROUND SYSTEM DiS 5s no substitute for an actual measurement of a “geeend system. Most measuring techniques are similar ‘See Gerefore, they have similar faults. ‘The techniques “Gis equipment with steady state de or low frequency s=ssveforms. Neither comes close to simulating the "Seamic surge conditions where inductive voltage drops veloped. Problems should not be incurred if good practices of low inductance are adhered to in the ‘and layout of the ground system, ‘These measur- meters can be used at a site to get an idea of the makeup. If no other ground conductors present, in or along a 100 foot path, a fall of (G stake) measurement can be setup before a rod under test is inserted into the ground. (See for theoretical ground resistance ) sites have a ground, but it is an unknown, The system is considered an unknown because it has been measured or if it was measured, it has iy changed. The resistivity varies through out ‘year because of seasonal moisture and temperature (The more moisture and the higher the ground wre, the lower the resistance.) A ground is like any other piece of equipment and normal smsintcnance must be performed to keep it in operating ‘ceedition. Unfortunately, it is out of sight and usually ‘ot of mind until a disaster strikes. Geound systems composed of copper and zine are y eaten away in acidic soils; yet are stable in the _gessence of alkaloids like concrete. Only aluminum is s=affected by acidic soils, but it is etched by alkaloids. ‘As mentioned previously, a simple test can be ‘secformed using a pool/spa water strip pH tester. ‘i addition to pH level, a soil's conductivity is deter- ‘ssved by its water and salt content. The more salts, the Jess water required to reach a specific conductivity. At ast 16% water content, by weight, is required for a seal to be conductive. Gypsum can be added to the soil ‘254 is better than bentonite. Gypsum absorbs and setains water and it doesn't shrink/pull away when ‘Geying like bentonite. Also, adding 5% by weight, of ‘epsom salts will further insure moisture retention and conductivity. tn 2 Like all partial conductors, the soil’s resistance value is measured in ohm-m or ohm-cm. A site's soil resistivity can be measured, (shown on page 25) by equally spac- ing four stakes (four stake method) and infusing current into the two outer stakes (C, and C>) while the two inner stakes (P, and P) measure the fall of potential (voltage). ‘There are several companies that make earth resistance testers operating at low frequency (90H2). ry Three stake method for measuring a ground rod or system (fall-of potential method). The three stake method, also known as the Fall-of- Potential Method, is shown and is used to measure the resistance of a ground rod or system. ‘This can be done on any four stake tester by tying P, and C; together. The initial spacing between electrodes P), C, and Cp for a simple clectrode would be approximately 100-fect, while for a ground system it would be approximately 1000-feet. The actual distance may be increased or decreased depending upon the size of the ground system being measured and the results of moving electrode Pp ‘The goal is to move electrode P at discrete intervals along a line between electrodes Pj, Cy and Cy and record/plot the voltage measurement, It is necessary tolocate the area of the curve where moving electrode P) has little or no affect on the measured voltage. The flat- tened area of the voltage curve is then converted using tables or formula supplied with the measuring instru- ‘ment fo the approximate ground system resistance Resign yourself to hammering in stakes to measure your ground system. The low frequencies used in most testers do not take into account any inductance which may exist in a ground system such as a rod penetrating a sandy layer, etc. The best way to determine the consistency of your underground soil layers is to perform a Fall-of-Potential measurement and log the readings for cach foot that a ground rod is driven Plotting it should approximate the Relative Earth Resistivity Curve shown here. Any large variation could mean water/clay or sand/gravel. With this knowledge, a better ground system can be designed for the RF properties of the lightning strike. RELATIVE EARTH RESISTIVITY CURVE $49 ‘RELATIVE TO FIRST ONE FOOT SECTION SURGE Z In IEEE Transactions on Broadcasting, Volume BC-25, No.l, March 1979, it was established that radials, together with rods, show a lower dynamic surge imped- ance under real lightning conditions than the resistance ‘measured at or near dc. This results from a lightning induced ground saturation causing localized arcing and creating a momentary low impedance path between ground masses. The effective area or size of the grounding system is thereby briefly increased. The 28 arcing is due to the fact that any ground system, no matter how good, will momentarily elevate above the global earth potential. This saturation may be due to a slow propagation of the surge and is measured as the velocity factor and time constant of the ground system, (See Chapter 12.) Obviously the larger the impulse, the more arcing and the lower the dynamic impedance. It has been shown, that the lower the measured impedance using the de or steady-state low frequency ac type instruments, the smaller the difference will be between the measured and the real dynamic impedance. GROUND PROPAGATION As in any medium, a dynamic pulse, like RF, will take time to propagate. This propagation time, together with resistive losses, will cause a differential step voltage to cexist in time between any two ground rods that are of different radial distances ftom the strike. With a ground rod tied to the base of a struck tower, the impulse will ideally propagate its step voltage out- wardly from this rod in ever-expanding circles, like @ pebble thrown into a pond. If the equipment hut has & separate ground rod and the power company and/or telephone company grounds are separate still, the dynamic step voltage will cause currents to flow to equalize these separate ground voltages. If the coax cable is the only path linking the equipment chassis with the tower ground, the surge will destroy circuitry while getting through to the telephone and power grounds.chapter 6 Putting it Together with Correct tying of Ground Rods together to form one ‘ground system, Bulkhead Panels 29 TYING IT ALL TOGETHER To prevent the multiple ground disaster from occurring, one grounding system must be formed which intercon- nects all other grounds together. A lightning strike pos- sesses a given amount of current and if there is a pe- rimeter ground loop around the building, the division of ‘current will be such that most of the lightning strike will bbe dispersed by the radials. This leaves a smaller ‘amount remaining to be handled by the perimeter ground around the equipment building. We use a pe- rimeter ring around the building because a loop will have a reduced amount of step voltage inside the loop. This is due to the repelling affect of like charge emanat- ing from all points on the loop. This hetps prevent cur- rent through conerete floors, protecting both equipment and personnel.ADVANTAGES DISADVANTAGES Low Liat voltage } 1) Coax must make tight bends. 2) Coax enters at flor level. Jp Line Protector Low L aiiat at tower | 1) Large L dit for inline protector unless large surface fea grounding conductor is used for building CGK and protector. 2) Sloped line will intercept tower mag fields Low L didt at building | 1) Coax must enter at floor level. 2) Sioped line wil intercept tower mag field. +) Enters building high. | Large straps cost more but are 2) Does net intercept | needed to reduce L. dict voltage. tower mag field.ytace is used tector. evel Hower are tage. EQUIPMENT STRESS Eeca with a "perfect" ground system, voltage stress Serng a lightning strike may still be experienced by the =qsipment if the coaxial cables are not brought to the seas of the tower, before the outside shield is grounded ‘ec #2 proper bulkhead panel is not installed with an “solated shield type protector (PolyPhaser's Isolated ‘Equipment protector), fe have already seen the tower's inductance and the ‘sssociated voltage drops. The tower connection at 15: S=t above the earth may appear to be a good grounding ‘pent An ohm meter might even show it to be a "good ground’, but it is really a poor, even dangerous ‘scunding location; dangerous to equipment and people. Fora 1 5/8-inch coax cable, it is virtually impossible to make as sharp a bend as is necessary to ground the “siicid at the tower base. Yet in the absence of other ‘grounding methods, itis essential to ground the shield at ‘Ses point to keep the shield voltage near zero. SF the coax slopes downward from an elevated point on ‘Se tower so it enters the equipment hut at or very near ‘scound level, another shield grounding kit should be in- ‘coxporated. 4% scither method is practical and the coax must enter ‘Se building significantly elevated above the earth, then ‘= large surface area grounding bulkhead panel should be used to ensure a good, low inductance grounding route BULKHEAD PANEL Balkhead panels have been used for many years. The ‘eitial reason was to provide the equipment hut with a structurally strong entry point for one or more cables. The bulkhead pane! was a rigid metal plate that covered ‘2 penetrating hole in the building's wall In installations where the coax lines must exit high on the tower, it is best to terminate coax lines at a bulkhead plate/coax protector and run a smaller more flexible line fom the coaxial protector to the equipment. Grounding the bulkhead plate has, in the past, consisted of 31 connecting a ground wire to this plate. A ground wire is very inductive - almost as inductive as 15-foot on the tower. A wire like this is ineffective, because of the inductance. A bulkhead plate grounded in this manner ‘will not strip surge current from the coax shield. > 119. ¢ Bulkhead Panel Coax Lines Ground Rod— ‘The Bulkhead Panel is grounded by cutting the pe- rimeter ground and bonding the ends to it. If the bulkhead plate is extended from the entrance hole along the exterior building surface and beneath the soil to the ground conductor, a low inductance interconnec- tion to the ground system can be made, (Because of its large surface area (skin effect) and large W/H ratio, it should be less inductive than an equal height on the tower.) If the coax cable is grounded to such a bulk- head plate, lightning surge current will be stripped from the cable shield. (To do this a grounding finger should be used. The attachment to the cable must be weather protected by a boot.) The ease of installation (weight) and cost of such a full length copper extension plate may be prohibitive. A cost effective variation is to substitute copper strap material (016" thick) for the thicker full length panel material going to ground. This makes it lighter, easier to install, and less expensive, ‘The strap should be affixed to the building with silicone and then covered or painted for camouflage and wind resistance. (The PolyPhaser bulkhead panel system comes in a variety of port configurations and has a built-in grounding kit that is protected by its UV stable -60) F to +300 F weather boot. It uses multiple 6" wide copper straps for grounding to a buried sandwich bar ‘that can be exothermically welded or lugged to the perimeter ground.)CENTER CONDUCTOR Shield currents can almost be eliminated with proper techniques. However, the center conductors surge current should also be eliminated before the current damages the equipment, A de blocked type lightning arrestor can prevent the center conductor's surge energy from reaching the equipment if itis mounted (grounded) ‘onto the bulkhead panel. The use of a PolyPhaser Iso- lated Equipment type protector will further prevent the sharing of surge energy present on the coax shield due to the L di/dt of the copper strap to the ground system. SUBPANEL To further protect and restrict access to the coax-to- arrestor connection, a subpanel is mounted/grounded to the bulkhead, The subpanel is attached so it protrudes from the main panel through the penetrating hole and creates a secondary surface on which arrestors are then mounted and grounded on removable circular plates. All connector matings are accessible from inside the building for tests and changes. If waveguide is used, the round plate is simply not installed, because a protec- tor is not needed. Proper grounding of the waveguide is, accomplished by the grounding finger, under the weather boot. The subpanel is deep enough (16-inches) to be used for concrete block construction and handles any cable(3), that doesn't arrive completely perpendicu- lar to the protector. The added depth allows for angle correction, ‘The best material to use for the bulkhead panel is 1/8- inch half hard C110 (solid copper). Only this hardness of copper can be properly tapped for screw threads. This copper weighs 5.81 pounds per square foot Mounting hardware used to join the subpanel to the bulkhead is 18-8 stainless steel For small to medium size sites, the bulkhead panel should be the central grounding point inside the building for single point grounding procedures. Holes can then be drilled into the U panel for bonding straps and cable lugs emanating from the racks of equipment. The bulk- hhead panel serves as the master ground window or ‘ground bus. Adjustable Grounding Fin in ‘Typical Coax Ajustable Buin ‘Grounding Strap Hand Aecass y Eatrance “area Por a o "Wide, 18.Long-|—» Cooper Straps Exotnermic 0 Lug Connections Sandwich Bare Grou 160° De pote Panels PolyPhaser oth Protectors _— Ground Strap Exiging Perimeter nd System Weather Boot over. Built-In Grounding StrapEnance Faris oun oval T simeter—! System T chapter 7 Inside Hut Grounding and Shielding Surely everyone has heard of the safety procedure that says to keep one hand in a pocket while working around high voltages. If the body does not complete a circuit, so current flows and danger is averted, To keep equipment safe in the event of a lightning strike, the same one-connection concept applies Single point grounding, as it is called, is a grounding ‘echnique that ties all the equipment in a building ther and grounds it at one common point. Imple~ menting this technique is quite easy for small installa- tions. vO PORTS Example: A simple repeater installation with a tele- phone interconnect. From the equipment’s perspective, there are three Input/Output (VO) ports: the coax, power line and telephone line. These V/O's can function as either a lightning source or sink. Lightning surge energy may originate from one 1/0 and exit another /O 33 causing circuit damage. It is impossible to ground an VO, so a surge protector must be provided for each, The surge protector's purpose is to divert, absorb and isolate the equipment from the surge. Whenever a surge exceeds a preset voltage, the surge protector diverts the surge to a ground sink, By installing a surge protector at cach V/O, it is possible to configure a grounding scheme that allows the equipment to survive a lightning strike‘a Por Bulkhead : Panel Assembly copper Ground Strap | Teloo Lines i eee | Scar Grouns exoroew oe” Interior Floor Level Bulkhead \gle Point Ground Preferred method of grounding VO protectors for coax. telephone line and power line. The bulkhead plate in- tum connects to the perimeter ground outside the equipment hut, A single point ground system would be created, if all the /O surge protectors were grounded/mounted onto a ‘bulkhead panel or onto a metal plate grounded to the ex- terior ground system with the equipment chassis also bonded here. Surge energy stripped from the lines by cach of the surge protectors is diverted to ground via a single path, Imagine each /O to be a hand or a foot. If a hand or a foot touch a high-voltage de source at a single point, no current would flow through the body, thus, no injury. (The surge current necessary to elevate the body up to this higher voltage might be felt.) The ‘body must therefore be insulated from everything else; no other path for current flow can exist. Likewise your equipment must be properly isolated using insulators which elevate the racks away from the conductive concrete floors. Some people may think this is unsafe for workers, because there will be a difference in potential between the floor and the equipment, This, i true, but the smart person will not be working at the site during a storm! By mounting the surge protectors ‘on the same bulkhead or metal plate which is grounded to a common ground, no surge current flows between the /O's and no voltage drop is created. Damage does not occur since the equipment chassis is also grounded to this same point, The surge protectors have a low im= pedance between them so no voltage drop can develop. PROTECTOR MOUNTING Current that is diverted by the protectors to. ground should be conducted by a path whose inductance is as low as possible, Ia grounded bulkhead panel, with its large surface area, low inductance, ground straps, is not used, the best place for mounting the surge protectors ‘would be on an inside, floor level, wall mounted plate, with a low inductance interconnection to the perimeter ‘ground. (This location is not as important in a high rise building room, unless the equipment is in a separate roof mounted hut.) ‘cops Protector gata 0 a. arco Tosaphone in, Braleaerie OPT. 1.12" Copper Strap Tote clouna Sytem Example of Single Point Ground without Bulkhead If the bulkhead plate was not installed at the time the equipment hut was built, an altemative grounding method may be used. Although the interconnect inductance of copper strap is greater, protectors may be mounted to a copper plate, which is connected by strap to the perimeter ground. Flat strap is the best conductor. It has maximum surface area, for skin effect and low inductance. If itis necessary to bend the conductor, such bends should be no sharper than an 8-inch radius. Strap actually has psge does ow im elop. ground > is as vith its is not tectors plate, gh rise parate em. chead 2 the nding ect ; may ed by itis ld be y has less inductance than wire for a given angle bend Mutual inductance, the cross coupling of the magnetic fields at the bend, is the reason for the added inductance of a bend. The distance from one side of the strap, when bent, is further away from the opposite side of the strap by the angle it makes, plus the width of the strap, The distance is greater so the mutual coupling is less. Also, the magnetic field susceptibility is maximum at its edges and itis similar to a dipole antenna, Therefore, it intercepts less tower magnetic fields if its flat side is oriented towards the tower. W CLIO OTD |A) Copper strap may be looked at as if made from an infinite number of infinitely small wires spaced infinitely close together B) the mag field of each wire C) is sown and they will add vectorially D) towards the edges. ‘No matter how low the protector path inductance may be, it still has some inductance (any conductor is also an inductor), Since the surge protectors divert current through this inductance, a voltage drop (L di/dt) is created. This voltage drop may cause problems for sensitive equipment. Control lines or balanced lines, for example, may become elevated above chassis ground. To ensure the equipment chassis is held to the same potential as the surge protectors, a low inductance connection between the equipment chassis and the bulkhead panel or protector panel is required. This conductor should have a lower inductance than the coaxial shield. SHIELD CURRENT FLOW If. Ufer ground with radials and a single point ground ‘are installed, have all the possible problems been liminated? NO! Surge current, however slight, could 35 still flow on the coax shield within the equipment hut, for two reasons: (1) The equipment chassis or rack, like your body, has the ability to accept a charge (capacitance), The chassis is elevated to the L difdt of the interconnection to the exterior ground. The current must flow to bring the chassis to this higher potential. This current flows along the coaxial shield (2) A large electromagnetic field is created by the pulsed current of the large lightning current surge. If the equipment is not in a shielded room (screen room) then the path from the equipment to the protector panel and to the ground system acts as an antenna, Since the ground system is a low-voltage, high-current point, the equipment end of the path, like the top of an antenna, is the high voltage point. This allows a portion of the pulse current induced by the electromagnetic field to flow along the coax cable shield This is an excellent reason to use PolyPhaser’s Isolated Equipment (IE) type protectors. They will help reduce the coax shield currents to almost zero MAGNETIC SHIELDING Lightning's high current means that the associated magnetic fields will radiate and cross couple to wire runs inside the equipment hut. Sites are usually designed to have a good 5 ohm ground system, but the building is placed right next to the tower with litte or ‘no magnetic shielding. The distance, between the tower and the hut, is usually kept small so the transmission lines are short. This places a heavier burden on your ground system to absorb and quickly conduct the strike energy away from the tower base. It also allows the ‘magnetic fields to enter your hut. Aluminum huts, like aluminum chassis, do little to attenuate low frequency magnetic fields, Concrete with steel mesh or rebar, which is ferrous, will show some attenuation. Steel shipping containers used as a hut, with either single or double walls, will act as a faraday shield for both radiated (plane wave) RF energy and ‘magnetic (H) fields. The containers also provide a good uniform ground for your equipment from anywhereinside the hut, You may not need steel electrical metal conduit (EMT) for shielding inside the hut, however, for other non-ferrous huts you should run all electrical and sensitive lines in separate EMT conduits. DISTANCE VS SHIELDING ‘The only altemative is to use distance. Magnetic fields drop off at one over the distance squared, To attenuate the strike's powerful H field from entering the circuit board and causing upset or damage, distance must be added. Distance will also add length to your transmis- sion lines which gives additional inductance (voltage drop), forcing more surge current down to the tower ground, This is unlike adding length/inductance with drip loops. The difference is the straight run from the tower to the hut is orthogonal to the magnetic field from the tower and will not pickup any additional surge (the drip loops will). Speaking of H field pickup, the orientation of the bulkhead panel strap is such that it is at minimum coupling and the strap(s) can give some small amount of shielding to equipment close to it and ‘opposite the tower, LATENT DAMAGE Attention to shielding is important because stress to electronic components can cause failure at a later date. ‘The US military has spent large sums of money to study ‘what has been termed "latent damage". Latent damage leads to premature MTBF (Mean Time Before Failure) of equipment. Lightning stress to high speed, small Jjunetion semiconductors, can lead to blue sky failures. Since the user does not have design control over the PC board layout, decoupling, trace length, proper VO protection, or the equipment enclosure, the only choice you have is to pick the manufacturer with the best ‘warranty and replacement service. LARGE SITE GROUNDING AND SHIELDING At very large sites (over 15m by 15m) where lightning, shielding is important and steel sheets can not be used to make a screen room, manufacturers sometimes provide an internal ground halo as an inexpensive altemative. This halo is both for personnel safety (grounding door jams and louvers) and to intercept the low radiated frequencies of lightning, however, it is not very effective. It often is mistakenly used for (multi- point) grounding of equipment racks GROUNDING EQUIPMENT CHASSIS Racks are commonly used to mount larger base station equipment and repeaters. Rack panels may be painted or the rails they are mounted to may become oxidized. ‘The paint and oxidation may have enough resistance to prevent the rails from being an adequate interconnecting conductor which ties all the rack mounted equipment together. Under non-screen room conditions and within high RF fields, such as those found at broadcast transmitter sites, the contact between the dissimilar metals of the bolts, rusty rack rails, and equipment panels can form "diodes". These "diodes" have been known to cause all kinds of intermodulation interference and audio rectfi- cation ‘A proper way to tie the equipment together is shown on page 37. The vertical ground bus is supported by insulators. Each piece of equipment is connected to the bus by a short strap. With this configuration, any “noise” created by poor joints and dissimilar metal contact within the rack is "shorted out" by the short loop. The short loop may be a resonant antenna near the frequency of the strong RF field from a nearby or co- located high-power broadcast. It may be resonant near ‘your operating frequency, or some other intermediate froquency used by your equipment. A grid dip meter may be used to determine whether the loop is resonant at a frequeney that could cause a problem. ‘The loop's resonant frequency can\be found with other methods. A spectrum analyzer may be link-coupled to the loop and observedito see where the noise:floor rises when the loop is opened and closed. The same tech- nique could be used with a service monitor tuned in.the AM mode to a quiet channel. (Note: Front-end over- loading could give false readings.)‘Esp equipment chassis in a rack is connected by “ex strap to a vertical grounding bus suspended by Bstctors. ‘GROUNDED SCREEN ROOMS Screen rooms work best for shielding equipment from ‘eb RF and electromagnetic pulse (EMP) fields associ- “seed with high-power transmitters, lightning strikes and Seehaltitude nuclear detonations. However, proper ‘ecthods of grounding the screen room itself shouldn't ‘be overlooked. The screen room manufacturer should ‘be able to detail the techniques that ensure maximum ‘sercen room effectiveness. MULTI-POINT GROUNDING Some microprocessor equipment manufacturers have opted multi-point grounding techniques in-an effort to exp noise and RF off the logic ground bus when ‘etcrconnecting to radio equipment. An example is ‘Sand in cellular switch locations. Since most of these ‘sc large installations, itis not practical to install a low 37 inductance interconnection (an inside halo) suspended in the air within the room (unless a screen room is used). ‘Therefore, independent interconnects to the perimeter ground are the answer (multi-point). Don't jump to this technique in haste. The problems that arise from not having a single point ground can be worse than the possible noise from the RF equipment. ‘While single point grounding is in most cases pre- ferred, large installations may make use of multi-point grounding to overcome the inductance of a ground interconnect conductor suspended in the air within the equipment hut. Careful design is required. PROBLEMS WITH MULTI-POINT GROUNDING Electrons take time to travel from one position to another (propagation time). For this reason, care must bbe taken in designing the perimeter ground system and in spacing the interconnects for a multi-point system. Serious trouble may result by too few interconnects to the perimeter ground, Remember, two hands (not one in the pocket) going to two different places can complete a circuit,For large equipment rooms, using many connections to the perimeter ground is a viable method. As a result, some surge current will enter the equipment room. However, with multiple paths to ground, the current is divided by the total number of paths. In this way the current is reduced to a harmless amount, ie., each L difdt is small enough so no damage occurs in the equipment. Ideally, each path from each piece of equipment to the perimeter should be of equal length. This means that for a typical site, the paths to the perimeter ground system should be interconnected to the inside halo about every two feet! This is rarely done and is why the halo has problems! It is easier to do a single point ground system than itis to install so many halo to perimeter interconnections. FAST PROPAGATION TIME ‘The multi-point ground design concept places more emphasis on the perimeter ground connections. Because the 1/0's are the only means by which current may enter the room, the lower its interconnect indue- tance to the perimeter ground by the numerous parallel paths, the smaller the L difdt voltage present. The propagation time of the ground system will be covered in detail later, but for now, the timing of the current in the earth around the hut can cause problems with a multi-point ground that is not present with a single point ground. Look at how the surge propagates in the following series of drawings. Note that the majority of the surge is diverted out and away from the building, Also note the time it takes to progress around the building. This is the reason why some currents traverse the parallel multi-point paths through the building in order to get to unsaturated perimeter ground locations. The race is between the speed of the perimeter versus the speed through the building. Normally if the ground system is any good, it will be faster than the slightly ‘more inductive path through the building. You can ‘wind up fighting yourself by adding more paths, making the voltage drop less, which then makes the propagation time faster through the building. This means more current will traverse through the building which is undesirable. ama Recommended site grounding system about to be hit by lightning Neglecting the coax currents, the strike energy moves outward from the tower base along the radial line.