January 2017
« Nov    



Lightning protection for WLAN links

More and more wireless LAN links are being installed between the transmitter and studio.  Often these links are used for network extension, remote control, site security, VIOP telephony, and sometimes even as a main STL.  These systems come in several flavors:

  • Moseley LAN link or similar system.  Operates on unlicensed 920 MHz (902-928 MHz) band.  Advantages: can use existing 900 MHz STL antennas, can work reliably over longer distances, transmitter and receiver located indoors.  Disadvantages: slow, expensive
  • ADTRAN TRACER or similar system with indoor tranceivers and coax fed antenna systems.  Operates on unlicensed or licensed WLAN frequencies.  Advantages: fast, transmitter and receiver located indoors, can be configured for Ethernet or T-1/E-1 ports.  Disadvantages; expensive
  • Ubiquiti Nano bridge or similar system where tranceiver is located in the antenna, the system is connected via category 5/6 cable with POE.  Operates on unlicensed or licensed WLAN frequencies.  Advantages; fast, relatively inexpensive.  Disadvantages; equipment located on tower, difficult to transition base insulator of series fed AM tower.
  • Ubiquiti Rocket or similar system where the antenna and tranceiver are separate, but the transciever is often located on the tower behind the antenna and fed with category 5/6 cable with POE.  Operates on unlicensed and licensed WLAN frequencies.

For the first two categories of WLAN equipment, standard lightning protection measures are usually adequate:

  • Good common point ground techniques
  • Ground the coaxial cable shield at the tower base and at the entrance to the building
  • Appropriate coaxial type transmission line surge suppressors
  • Ferrite toroids on ethernet and power connections

For the second two types of WLAN equipment, special attention is need with the ethernet cable goes between the tower and POE injector or switch.  Shielded, UV resistant cable is a requirement.  On an AM tower, the shielded cable must also be run inside a metal conduit.  Due to the skin effect, the metal conduit will keep most of the RF away from the ethernet cable.  Crossing a base insulator of a series excited tower presents a special problem.

The best way to get across the base insulator of a series excited tower is to use fiber.  This precludes the use of POE which means that AC power will be needed up on the tower to power the radio and fiber converter.  This my not be a huge problem if the tower is lit and the incandescent lighting system can be upgraded to LEDs.  A small NEMA 4 enclosure can house the fiber converter and POE injector to run the WLAN radio.  Some shorter AM towers are no longer lit.

Another possible method would be to fabricate an RF choke out of copper tubing.  This is the same idea as a tower lighting choke or a sample system that uses tower mounted loops.  I would not recommend this for power levels over 10 KW or on towers that are over 160 electrical degrees tall.  Basically, some 3/8 or 1/2 inch copper tubing can be wound into a coil through which a shielded ethernet cable can be run.  Twenty to twenty five turns, 12 inches in diameter will work for the upper part of the band.  For the lower part, the coil diameter should be 24 inches.

In all cases where CAT 5 or 6 cable is used on a tower, it must be shielded and the proper shielded connectors must be used.  In addition, whatever is injecting power into the cable, ether POE injector or POE switch must be very well grounded.  The connector on the shielded Cat5 or 6 cable must be properly applied to ensure the shield is grounded.  A good video from Ubiquiti, which makes TOUGHCable, on application of connectors to shielded Cat5 cable is here:

In addition to that, some type of surge suppressor at the base of the tower is also needed. Tramstector makes several products to protect low voltage data circuits.

Transtector APLU 1101 series dataline protector

Transtector APLU 1101 series dataline protector

These units are very well made and designed to mount to a tower leg. They come with clamps and ground conductor designed to bolt to a standard copper ground buss bar.

Transtector APLU 1101 series dataline protector

Transtector APLU 1101 series dataline protector

There are various models designed to pass POE or even 90 VDC ring voltage.

Transtector APLU 1101 series dataline protector

Transtector APLU 1101 series dataline protector

This model is for POE. The circuit seems to consist mostly of TVS diodes clamping the various data conductors.

As more and more of these systems are installed and become a part of critical infrastructure, more thought needs to be given to lightning protection, redundancy and disaster recovery in the event of equipment failure.

Series surge suppressor

Radio facilities, particularly mountain top transmitter sites, are prone to power transients. The causes can be varied, but most often, lightning is the culprit.  Long power transmission lines to the site are vulnerable to direct strikes and EMF induced spikes from nearby strikes.  Other issues, such as switching transients, load fluctuations, and malfunctioning equipment can lead “clear weather” outages.  Of course, the best way to deal with such things is prevention.

Power line surge suppressors have been around for quite some time.  They usually take the form of a MOV (Metal Oxide Varistor) connected between the hot leg and neutral or ground.  There are a few differences in designs, however.  Typically, most facilities employ a parallel surge suppressor.  That normally take to form of an enclosure hung next to the main power panel with a group of MOV modules in it.  The MOVs are feed from a circuit breaker in the panel.  Like this:

LEA parallel surge suppressor

LEA parallel or shunt surge suppressor

This is an LEA three phase 208 volt shunt surge suppression unit, which has MOVs between all phases to ground and each other.  That is connected in parallel to the electrical service with the circuit breaker disconnect.  These function well enough, provided there is a good bit of series inductance before the unit and also, preferably after.  The series inductance can come from many sources, including long secondary leads from the utility company transformer or electrical conductors enclosed in metal conduit, particularly rigid (verses EMT, or FMC) metal conduit.  The inductance adds a bit of resistance to the transient voltages, which come in higher than 50 or 60 Hz AC waveform.

A better method of transient protection is the Series Surge Suppressor.  These units are installed in line with the incoming service and include an inductor to add the required series resistance coupled with MOVs and capacitors.  Most series surge suppressor also filter out harmonics and RF by design, something desirable particularly at a transmitter site.  Series surge suppressors look like this:

LEA DYNA systems series surge protector

LEA DYNA systems series surge protector

This is a LEA three phase 240 volt unit.  As in the other example, all phases have MOVs to neutral and each other.   There are MOVs and capacitors on the line and load side of this unit (line side is the bottom of the inductor).  A basic schematic looks like this:

Series surge suppressor basic schematic

Series surge suppressor basic schematic

A few things to note; MOVs have a short circuit failure mode and must be fused to protect the incoming line from shorts to ground.  MOVs also deteriorate with age, the more they fire, the lower the breakdown voltage becomes.  Eventually, the will begin to conduct current at all times and heat up, thus they should also be thermally fused.  MOVs that are not properly protected from over current or over temperature conditions have the alarming capacity to explode and/or catch on fire.  From experience, this is something to be avoided.  Matched MOVs can be paralleled to increase current handling capacity.

The inductor is in the 100 µH range, which adds almost no inductive reactance at 60 Hz.  However, it becomes more resistive as the frequency goes up.  Most transients, especially lightning, happen at many times the 60 Hz fundamental frequency used in power distribution (50 Hz elsewhere unless airborne, then it may be 400 Hz).

Capacitors are in the 1-10 mF range and rated for 1 KV or greater as a safety factor.  The net effect of adding capacitance is to create a low pass filter.  Hypothetically speaking, of course, playing around with the capacitance values may net a better lowpass filter.  For example, at 100 uH and 5 mF, the cutoff frequency is 225 Hz, or below the fourth harmonic.  Care must be taken not to affect or distort the 60 Hz wave form or all sorts of bad things will happen, especially to switching power supplies.

These units also need have a bypass method installed.  If one of the MOV modules needs to be replaced, power to the unit has to be secured.  This can be done by connecting it to the AC mains before any generator transfer switch.  That way, the main power can be secured and the site can run on generator power while the maintenance on the surge suppression unit is taking place.

The Polyphaser IS-PT50HN-B

I found this on the floor at an old transmitter site:

Polyphaser IS-PT50HN-B DC block surge suppressor

Polyphaser IS-PT50HN-B DC block surge suppressor

Since it appears to be discarded, I ignored the dire warnings and opened it up to look inside:

Polyphaser IS-PT50HN-B DC block surge suppressor

Polyphaser IS-PT50HN-B DC block surge suppressor

This is is a DC blocked lightning surge suppressor designed for 890-980 MHz, 750 watts maximum.  The two parallel wires represent a capacitor, coupling the radio to the antenna, the inductor acts as an RF block to the gas discharge tubes.  The design is such that the inductor acts to block the normal in use radio frequencies but will allow the 10-30 KHz lightning pulse to pass to the gas discharge tubes thence to ground.  The inductor and gas discharge tubes are on the antenna side of the unit.  I measured these units with a DVM and they all appear to be good.

My only comment on this unit is that there is no effort to maintain the transmission line impedance.  At the upper end of the UHF spectrum, this can lead to return loss and wasted power.  For a receive application, it may not be so bad, but for a transmitter, I would rather use something else.

For lower VHF frequencies, something like this can be DIY fabricated with minimal expense and effort.  The case must be bonded to the station ground.

Lightning season

Here in the northeast, there are seasonal various in the types of weather phenomena encountered.  Blizzards in the winter, severe thunderstorms and the occasional tornado in the summer, at least that is the way it normally happens.  This year, we have already had two thunderstorms and a stretch of unusually warm weather.  My highly advanced personal weather prognostication technique consists of looking at trends, and the trend thus far this year is warmer with more storms.

Weather Radar, thunderstorm line

Weather Radar, thunderstorm line

When the weather RADAR looks like this, it is too late.

To that end, it is time to go around and check all of the grounding and lightning suppression methods at various transmitter sites and studios.  I would rather spend a few minutes extra now than get called out in the middle of the night for an off air emergency related to a lightning strike.

Proper grounding of all equipment, RF cables and electrical service entrances are the minimum standard for transmitter sites.  Proper grounding means a common point grounding system connected to one ground potential.

To that end, all coaxial cables that enter the building need to have their outer shields bonded to the site grounding system at the base of the tower and the entrance of the building.  With an FM station where the antenna is mounted at the top of a tall tower, the coaxial cable outer jacket acts as in insulator along the length of the tower.  A lightning strike on the tower will induce a very high potential on the outer conductor of an ungrounded transmission line.  After entering the building, the lightning surge will find the next path to ground, which will likely be a coax switch or the transmitter cabinet.  Neither of those two outcomes is desired.

Thus, it was time to ground the transmission lines at WRKI, the FM transmitter we moved last January.

3 inch coaxial cable grounding kit

3 inch coaxial cable grounding kit

Fortunately, Andrew, Cablewave, Dielectric and others make grounding kits for various size coaxial cables. They are very easy to apply and make a solid connection between the outer conductor and the site ground.

3 inch coaxial cable grounding kit

3 inch coaxial cable grounding kit

The kit contains a copper band bonded to a ground wire, stainless steel clamp, water proofing, tape and a pair of bolts.

3 inch coaxial cable properly grounded

3 inch coaxial cable properly grounded

The concept of transmitter site grounding is pretty simple and inexpensive to implement.  Thus, it is surprising to me how many transmitter sites, especially older sites, that do not have adequate grounding.  That is an accident waiting to happen.

For more on transmitter site grounding, check Nautel’s publication (.pdf) “Recommendations for Transmitter Site Preparation.”

The melted ground wire

I found this on one of the guy wire anchor points for a 400 foot tower:

#2 solid copper wire burned open by lightning strike

#2 solid copper wire burned open by lightning strike

Had to be a pretty big hit to burn open a #2 wire. This is on one of six guy anchor points for the tower. The ground wire is U bolted to each guy wire before the turnbuckle and then goes to ground. This was noted between the last guy wire and the ground rod.

It is important to find and fix these things, as the next lightning strike on this tower would have a less than ideal path to ground at the guy anchor points, forcing the current to flow through other parts of the transmitter site, possibly through the transmitter itself, to ground.

I generally try to do a brief inspection of towers, guy anchors, lighting, painting and a general walk around the property twice a year.  That helps prevent surprises like “Oh my goodness, the guy wires are rusting through,” or “Hey, did you know there is an illegal “hemp” farm on your property?”  Well, no officer, I don’t know anything about that…

Burning Telegraph Wires and other Irregularities

By now, you have heard of the great solar flare of 1859 that set telegraph wires afire across the US and Europe.  This phenomena was due to a large electro-magnetic pulse from the earth’s magnetosphere interacting with the particles from a Coronal Mass Ejection (CME) caused by the solar flare.  The long stretches of wire suspended above the earth acted like a large generator winding cutting through the magnetic field, which in turn caused voltage (Electro-magnetic force or EMF), which in turn, caused the fires.

In 1859 our understanding of electricity and magnetics is not what it is today, thus, the fires were likely caused by over loaded conductors with out means to bypass excess electrical energy to ground.

Schematic of Earth's magnetosphere, courtesy NASA

Schematic of Earth's magnetosphere, courtesy NASA

Fast forward to today.  We are in the upswing of solar cycle 24, which is expected to peak in May of 2013.  NASA predicts that this will be lowest sun spot peak since 1907.  That does not necessarily mean the coast is clear.  Over the last 11 years since solar cycle 23 peaked, computers and electronic automation have proliferated exponentially, becoming the norm.  Things like GPS not only guide clueless travelers where to turn, but also sync up all those cellular telephone transmitters with timing signals.  IP networks, SCADA, telephone networks and so on all run on some form of CPU.  Newer Energy Star appliances like toasters and refrigerators also have CPUs.   That technology has yet to experience a large electro-magnetic pulse (EMP) in the real world.  Things could indeed get hairy if a moderate to large class X solar flare generated a CME that was polarized correctly to interact with Earth’s magnetic field and cause damage at ground level.

Solar flares and CME are slow moving events, with 1-2 days warning before the effects of a CME reach earth.  One can stay apprised of solar flares and other solar activity by subscribing to NOAA Space Weather Prediction Center’s email service.

HEMP Mechanism

HEMP Mechanism for 400 kM high altitude burst

Of greater concern is other sources of EMP like high altitude nuclear explosions (HEMP).  Those types of events, while rare, can happen.  The good news is, the defense mechanisms for solar flare, high altitude nuclear burst and lightning induced EMP are the same.  They are effective grounding, shielding, filtering and surge suppression.1  Of the three EMP scenarios, high altitude nuclear burst has the tightest design spec, so creating a building that incorporates the ideas in MIL-STD-188-125-1 specifications is a good start.

The question becomes, is all of this really necessary.  It depends on how important it is to the radio station ownership to remain on the air during such an event.  Based on historical information and global geopolitics, the probabilities of such an occurrence are:

  • Lightning strike – 1:1 Any radio station that has a tall structure, particularly a steel tower, will get struck by lightning, perhaps several times per year depending on the region.
  • Large Class X solar flare resulting in damaging CME – 1:21 Since 1859, there have been seven solar flare events that have disrupted communications or power systems on earth.  This is a bit misleading since 6 of the 7 events have occurred in the last 22 years, making the real probability more like 1:3.6
  • High Altitude EMP – 1:30 Based on seventeen  high altitude tests carried out by the US and USSR in 1962, the growing nuclear proliferation and a June 2005 Reuters article “Experts warn of substantial risk of WMD attack” in which the author stipulates a thirty percent risk of a nuclear attack of any type in the period of 2010 to 2015.

EMP Theory

High altitude nuclear burst EMP has three components; the fast component (20/550 ns pulse) is an electromagnetic shock-wave, the medium-speed component (1.5/5000 μs pulse), and the slow component (0.2/25 s pulse) resulting from the expansion of the explosion’s fireball in the Earth’s magnetic field.1  Compare that to a lightning strike, which typically has a 1.8 µs rise time.  That means the first pulse frequency is from about 72 to 200 MHz, second pulse frequency is from about 800 Hz to 2.5 MHz and third pulse is basically DC and effects mostly long wires.  Thus, any shielding, grounding and suppression needs to consider the highest frequency down to about 10 KHz.2

The rise time to frequency comparison is an important consideration in the design and construction of grounding systems.  Grounds need to present the least amount of inductance possible.  This means using solid, not stranded wire, keeping bends to a minimum and where required, use long radiuses.  Bond all junctions by welding, exothermic welding or soldering.  Use a single point ground system.  Zone grounding should also be employed.  The definition of zone grounding is concentric grounding areas connected to each other by a single low inductance ground conductor.2  The idea of isolating grounds by use of a ferrite suppressor may seem odd, however, if there is a separate RF ground, tied together with the building ground using wide copper strap laid on the floor, this will minimize ground system “reception” of incoming EMP.2,3

Zone grounding diagram

Zone grounding diagram

Shielding means surrounding the protected equipment with a conductive material such as copper plate, aluminum plate, copper mesh, aluminum mesh, brass mesh, steel plate or steal mesh.1,2  There are advantages and disadvantages to each.  Seams should be welded to prevent “leaks.”  Doors need to have finger stock or metal compression gaskets to ensure proper sealing of opening.  Other openings for ventilation, cable ingress, etc. need to be “significantly” less than one wave length.  MIL-STD-188-1 gives the allowable opening size of 10 x 10 cm or 3.9 x 3.9 inches. If openings in the shield become greater than approximately wavelength/6 meters (at 250 Mhz, about 8 inches), significant fields can penetrate to the interior.2

Suppression deals with those connections to the areas outside of the facility.  These include incoming electrical service, data service and RF transmission lines to and from antennas.  Since the fast component of the HEMP falls within the VHF spectrum, FM broadcast installations are particularly vulnerable.  Suppression devices for incoming AC power are readily available from commercial sources and are well proven.  LEA makes a series surge suppressor that uses a combination of fast acting silicone diodes, MOVs and an LC filter made up of series inductance and parallel capacitance to ground.  The LEA DYNA family series surge protectors have a system response time of less than one nano second and are tested to greater than 1,000 operations.4  The response time depends on a good, non-inductive ground connection.

LEA DYNA systems series surge protector

LEA DYNA systems series surge protector

Suppression for incoming RF and data cables is more difficult because the normal operating frequencies fall within the HEMP rise time frequency.  Incoming data at a transmitter site usually consists of a DS-1 circuit however, larger capacity circuits are sometimes used.  Fiber optic cables are immune from HEMP as they have no metal conductors.  Copper data lines must have a data line surge suppressor between the TELCO demark and the CSU.

RF cables must have their shields ground to the zone 0 ground, then go through a ferrite toroid to add inductance to the outer shield and isolate it from the zone 1 ground.  After the ferrite toroid, a gas discharge type inline surge suppressor should be used.  These come in a variety of configurations, frequency band and power levels.  It is best to keep the suppressor rating as close to the peak carrier power as possible, affording the most protection to the transmission equipment.

Design and implementation of EMP hardened facilities

Of the four strategies for mitigating HEMP; Grounding, Suppression, filtering and shielding, shielding is the hardest and most expensive to implement.  Good grounding should be included in any good radio station design same as suppression and filtering.

  1. Grounding.  Grounding for a transmitter site must include an outside ring ground around the periphery of the building.2  This is bonded to several ground rods installed at regular intervals.  The ground is brought into the building and all coax shields, electrical service entrance, TELCO equipment, suppression equipment and safety grounds are connected to it.  This forms the zone 0 ground.  One conductor then goes to the zone 1 ground which is the transmitters and racks.  Any other conductors at any potential that go from zone 0 to zone 1 are bypassed at EMP frequencies by use of ferrite toroids or other high mu ferrous metal filters.  Inductors may need to be bonded to ground to prevent saturation.  At studio locations, the building electrical safety ground should be evaluated for adequacy.  Additional grounding may need to be installed depending on effectiveness of existing ground. Any outside antennas, supporting structures, satellite dishes and generators need to be bonded together and grounded to the building electrical grounding system.  Studios and engineering rack rooms need to be bonded to the ground using star topology.  Facilities that are not adequately grounded should be retrofitted.
  2. Suppression and filtering.  Good surge suppression and filtering should be a part of all transmitter and studio site designs.  Hanging a few MOV’s off of the service panel is not enough to prevent damage to a facility.  All incoming lines from the street; electric, telephone, and cable need to have surge suppression connected and be bonded to a low inductance path to ground.  The only exception to this is fiber optic, which is immune to the effects of EMP. Additional layers of filtering for sensitive, mission critical computer systems such as FERUPS, shielded category wiring that is properly installed, etc.  Facilities that do not have adequate suppression can be retrofitted.
  3. Shielding.  Shielding is the most expensive, time consuming and difficult to install correctly.  The High Altitude nuclear test Starfish Prime in July of 1962 produced a field of 5600 V/m in Honolulu, some 1300 KM away from the blast.  Building a shielded structure against those intense magnetic and electrical fields is very difficult.  Attenuating the field through layers of shielding is the most effective means provided the distance between the shields is wavelength/6 or more (about 27 inches at 72 MHz) to prevent coupling.2  For example,  using a concrete structure with steel mesh creates 35-40 dB of attenuation in zone 0, a well designed transmitter with good RF shielding in it’s cabinet design creates a shield for zone 1.  Equipment racks can also be used to create shielding zones by using copper or brass mesh with good metal to metal contact around front and back doors.  At studio locations, engineering rack rooms should have copper or brass mesh embedded in the wall structure to create a shield.  This will create a safe area to locate computer file servers, routers, switches, STL gear, satellite receivers and the like equipment.  Layered shielding with use of metal, gasketed door will improve shield performance.  Retrofitting shielding is more difficult to accomplish than grounding and suppression.  It is best done in new construction.  There are many different ways to accomplish even moderate shielding, which may serve well for lightning and solar flare induced EMP.
From personal experience, investing an extra $10-20K in grounding and suppression at a lightning prone transmitter site in Florida solved all of the issues at that site.  Prior to installing the ring ground and bonding, the transmitter was knocked off the air several times per year.  Since the work was done in 2005, there has not been one lightning related outage at that site.

1. Protection Technology Group, System Approach to EMP Mitigation, February 2011
2. US Army Corps of Engineers, Engineering and Design – Electromagnetic Pulse (EMP) and Tempest Protection for Facilities EP 1110-3-2, December 31, 1990
4. Protection Technology Group, Modular Hybrid Series Connected Surge Protection Device LEA DS21 data sheet, 2010

Lightning Season

The rumble of thunder this morning let me know that lightning season is upon us here in the Northeast and likely the rest of the country as well.  I used to enjoy the odd summer thunderstorm, especially the late afternoon pop ups that cool off a hot summer’s day.  Now whenever I see lightning or hear thunder I wonder if the phone is going to ring.  Chances are good that it will not, as I invested many hours of my time and my previous employer’s money into lightning protection at the transmitter sites.

upper atmosphere lightning depiction

upper atmosphere lightning depiction

I go on the assumption that all tall steel towers will get struck, often times repeatedly, during any particular electrical storm.  Back in the day, I took a course by Polyphaser called “The grounds for lightning and EMP protection.”  It was a great primer on how to ground and bond equipment at a transmitter site to eliminate current flow, which is the cause of all EMP and lightning inducted failures. When I was in military communications, no expense was sparred as they took uptime very seriously.  Any downtime was a personal affront to the commanding officer of the unit in question.

Lightning is DC however, it behaves more like 10 KHz – 2 MHz AC.  Therefore, lightning and EMP grounding systems need to be designed and installed to accommodate DC through 10 KHz AC voltages.  This is easily done by choosing the correct conductors, ground buss bars and bonding systems.

The other path lighting takes into a transmitter is through the AC mains.  Utility company high voltage primary feeds act like large antennas for lightning induced EMP.  Fortunately, much of that is filtered out by the step down transformers just before the building service entrance.  It is still possible, however, for some impulse voltage to make it though the transformers and into the service entrance panel.  On older tube type transmitter, this could often damage the plate voltage power supply because of the voltage multiplication factor of the plate transformer.  Often times, the transformer secondary would have “holes” punched through the insulation to ground.  This is an expensive and time consuming repair.

I would conservatively estimate that for every $10.00 spent on lightning protection, $1,000.00 dollars worth of damage and downtime is saved.  Overall, a pretty good return on investment.

The basics for lightning ground bonding are thus:

  • Use the lowest inductance wire possible.  The industry standard is #2 solid copper, however, if bonded properly, there will be very little current flow inside the transmitter building, so if #2 is not available, then any solid wire up to #8 will work.  Tower ground bonding should be as heavy as possible.
  • Ground all guy anchor points, bond all guy wires together and to the same ground rod or ground rod system.
  • Keep the bonding conductors as straight as possible, bends should be long sweeping turns to minimize series inductance.
  • All metal equipment should be bonded, no rack, telco demark, electrical panel, dummy load, bulkhead entrance grounding buss, combiner, door frame, etc should be left unbonded.
  • All coax outer shields should be grounded where it comes into the building.
  • All coax cables should go through a toroid before being connected to the transmitter.
  • All outdoor bonding connections should be exothermically (CAD welded) bonded to ground rods.
  • All grounding must go to a common ground point, AKA star grounding point.  No individual ground points should be allowed in the building.
  • Multiple ground rods installed around the outside perimeter of the building as deeply as possible.  Some mountain top transmitter sites may require special grounding material (Bentonite) and or to have a ground well drilled.  Ground conductors should have as much surface area contact with earth as possible.

The whole idea is to present a low resistance ground path and keep all of the equipment at the same potential to minimize current flow between equipment.

For the electrical building service entrance, a series surge protector installed before the service panel is the best method.  Several are made and they need to be sized for the building service.  A fall back is a parallel surge protector will provide some protection.  On the AC mains connections, any series inductance that can be added to increase resistance to the lightning pulse is good.   All AC mains connections to the transmitter should go through a toroid before they are connected to the transmitter.  This is a good idea for remote control and mod monitor wiring as well.


Lightning Damage

It is that time of the year again, at least in the northern hemisphere, for thunderstorms.  I am a big proponent of grounding everything, there is simply no such thing as too much grounding.  I took a course when I was in the military given by Polyphaser in which grounding for lightning protection and EMP was emphasized.  It was very interesting in several respects.

One commonly held belief is that when lightning strikes an object, the ground immediately absorbs all of the charge.  That is not true in most cases due to ground resistances.  Eventually, the ground will absorb the charge but it can take several seconds to do this, especially with a big strike.  Equipment is damaged by current flow, therefore, every effort must be made to keep all of the equipment at the same potential, even if that potential is 10KV.  That is where a single point ground buss comes in.  Bonding every piece of equipment to a common ground buss ensures that no one device is at a lower potential while the charge dissipation is occurring.

The second misunderstanding about lightning is that it is DC voltage.  That is true, however, a lightning strike has an extremely fast rise time, on the order of 30 microseconds.  That makes it behave more like AC voltage around 10 KHz.  Therefore, ground buss wires need to have a minimum inductance.  Solid #2 wire is best, keeping it as straight as possible and using long sweeping turns where needed.  All bonds should be exothermically welded (CAD weld).

Ground system installed at WKZY, WHHZ and WDVH, Trenton Florida

Ground system installed at WKZY, WHHZ and WDVH-FM transmitter site in Trenton, Florida.  Central Florida is the lightning capital of the US.  Prior to doing this work, the Harris FM25K transmitter was knocked off the air at least once a month.  Since this was installed in 2005, they have had zero lightning related damage.  The ground rods are 20 feet long, driven down into the water table, spaced 20-30 feet apart.

All coax shields and metal conduits that come into the building should be bonded to the ground system where they leave the tower and where they enter the building.  At most tower sites, I install a ground ring around the outside of the building with rods every 20 feet or so.  From that ring, 5 to 6 radials outward 40 feet with ground rods every twenty feet works well.  I also install 5 to 6 radial out from the tower base with the same configuration.  The tower and building grounds are bonded together.  This is important because when the tower gets hit, the ground will quickly become electrically saturated.  If the building and the equipment is inside is at a different potential, current will begin to flow toward the lower potential, thus damaging gear.

All Coax, control and AC cables in and out of sensitive equipment should have ferrite toriods on them.  Transmitter manufactures normally supply these with new solid state transmitters, as MOSFETS are particularly sensitive to lightning damage.

Lightning damage to rack mounted equipment

This is a Potomac Instruments AM-19 directional antenna monitor.  It was damaged by a lightning strike two weeks ago on the WBNR tower in Beacon, NY.  The case arced to the rack it was mounted in.  This was a large strike, as several components in the phasor control circuit were also damaged.  The fact that this arced means that somehow the sample lines are not attached to the single point ground for this site, which needs to be corrected.

Insulated AM towers present special design problems when it comes to lightning protection.  Generally speaking, tower arc gaps should be set so there are side by side and there is no arcing on positive modulation peaks.  Depending on power levels, this can be anywhere from 1/2 inch to 2 inches.  Tower impedance also plays a roll in setting arc gaps.  The final link between the ATU and tower should have several turns in it.  The idea is to make that path a higher impedance path for the lightning, causing it to dissipate through the arc gaps.  Incoming transmission lines from the towers should be bonded to a copper buss bar at the entrance to the building.  All of this grounding needs to be tied to the RF ground at the base of the tower.

Arial phone cables can act like large lightning antennas for strokes several miles away.  It is very important that the cable shield and the cable termination device is bonded to the building ground buss.  I have seen installations where the TELCO tech pounds in a separate ground rod outside and connects the TELCO equipment to that.  That defeats the concept of single point grounds and should be fixed ASAP.

Electrical services entrances also can act like big lightning antennas.  Normally, pole mounted transformers will filter some of this energy out.  Internal electrical distribution systems can also add impedance, thus act as inadvertent filters for lightning.  In most mountain top transmitter sites, however, some type of power line surge protection is needed.

LEA series surge protector

Inside view of LEA surge suppressor

There are two types, series and parallel.  Parallel types are the least expensive and least intensive to install.  They are usually found mounted next to or on the service panel and fed with their own breakers.  They usually have some type of MOV or similar device that acts as a crowbar across the AC mains, conducting spikes to ground.  Series types go in between the service entrance and the main panel.  They include a large inductor designed to force spikes off into shunts.  A series type protector offers more complete protection than a parallel.

Delta Current Sample Toroid

Another example from my blown up shit collection, artifacts division:

Delta TCT-1HV current sample toroid destroyed by lightning

Delta TCT-1HV current sample toroid destroyed by lightning

This is a Delta TCT-1HV current sample toroid that was pretty well destroyed during a thunderstorm.  I mounted it on a piece of plexi-glass because I think it looks cool.  This unit was installed at the base of the WGY transmitting tower.  One June evening, I received a call from the station operator (back when they had live operators) that the air signal sounded kind of “funny.”  So I turned on the radio and sure enough, if one thinks a radio station that sounds like a motor boat is funny, then, why yes indeed, it did sound funny.

Since I only lived a few miles away from the site, I jumped in the trusty truck and headed over.  Upon arrival, I found the MW50B on the air at full power, with the carrier power swinging wildly from 20-90 KW with modulation.  Hmmmm, bad power supply?  Turned the transmitter off and tried to place the backup transmitter on the air.  Now the old Gates BC5P had never been super reliable in the first place, but it was odd that it would not even run at all.

Then I had a hunch, lets walk out to the tower I said to my assistant who had showed up to help.  When we got to the ATU building it was filled with blue smoke.  Ah ha!  Somebody let the magic smoke out of one of the components!  I was expecting a capacitor blown in half but was surprised to fine the copper tubing that connected the ATU to the tower melted in half.  Lightning must have caused an arc between the tubing and the toroid and for some reason the transmitter kept on running while it was arcing.  The copper tubing in the picture with toroid is only missing about six inches, the way the system was mounted at the tower base, fourteen inches of copper tubing was missing, or rather melted into a puddle on the bottom of the ATU.

I quickly found another piece of 1/2 inch copper, cut it to length and flattened out the ends with a hammer and drilled mounting holes.   Luckily I was able to get everything back in order quickly and the station returned to the air about an hour or so after it went off.

Everything has a cause.  Investigation showed that the VSWR circuit on the MW50 had been disconnected from the directional coupler.  The lead was un-soldered and taped off, so it was quite intentional.  I spoke briefly with two of the three prior engineers that had serviced the MW50 over the years, they both blamed the other one.  I surmise this; The WGY tower was prone to lightning strikes because of it’s height.  Even if the tower was not directly struck by lightning, often times the guy wires would arc across the insulators, causing the MW50 to momentarily interrupt the PDM signal and drop the carrier for about a second.  Some programming people at the station did not like this, it sounded bad on the air, so one of those guys undid the VSWR circuit and voila! No more momentary outages during a thunderstorm! Brilliant!  Except for the 60-90 minute outage one night…

Sometimes it is better to tell the program directors that their idea is not good, then move on.

Lightning strikes

For about 4 years, I lived in a house next to a four hundred foot radio tower.  Although I never actually saw lightning strike the tower, I heard it several times.  Like everything else, after a while, you get used to it.

This is a video of lightning hitting the WSIX STL tower in Nashville, TN. The camera work is a little unsteady, the strike occurs around the 1:36 mark:

Yep, that is what it is like.


A pessimist sees the glass as half empty. An optimist sees the glass as half full. The engineer sees the glass as twice the size it needs to be.

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.
~1st amendment to the United States Constitution

Any society that would give up a little liberty to gain a little security will deserve neither and lose both.
~Benjamin Franklin

The individual has always had to struggle to keep from being overwhelmed by the tribe. To be your own man is hard business. If you try it, you will be lonely often, and sometimes frightened. But no price is too high to pay for the privilege of owning yourself.
~Rudyard Kipling

Everyone has the right to freedom of opinion and expression; this right includes the freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers
~Universal Declaration Of Human Rights, Article 19

...radio was discovered, and not invented, and that these frequencies and principles were always in existence long before man was aware of them. Therefore, no one owns them. They are there as free as sunlight, which is a higher frequency form of the same energy.
~Alan Weiner

Free counters!