Grounding satellite systems thru coax shield

Status
Not open for further replies.

BarryO

Senior Member
Location
Bend, OR
Occupation
Electrical engineer (retired)
Hello,
I was hoping that the people here, being close to the NEC-writing process, could address a topic that?s become something of a big controversy among some satellite systems installers, electricians, and engineers. It pertains to the installation standards for a new satellite broadband service from a company named WildBlue (www.wildblue.com).

The exterior components of the Wildblue system consist of a parabolic dish antenna mounted on a short mast, with some active transmit/receiver electronics in a small enclosure termed a Transmit/Receive Integrated Assembly, or TRIA. The TRIA is connected to the modem inside the house by a double run of coaxial cable, over which the TRIA also receives DC power.

Unlike other DBS and broadband satellite systems, the Installation Manual for the Wildblue system does not specify that the dish & mast be grounded with a separate grounding conductor. Instead, it specifies that the dish be explicitly bonded to the TRIA with a #10 Cu wire. It goes on to comment that the TRIA is grounded through the coax shields.

This lack of an explicit grounding conductor for the mast in accordance with 810-15 has caused some to maintain that the Wildblue installation manual is not in compliance with the NEC. Reading ?810 carefully, I don?t believe that is necessarily the case.

Article 810 covers antenna systems for radio, television, and amateur radio. While it doesn't explicitly mention dishes for broadband access, to the extent that any local enforcement entity cares about antenna compliance with the NEC, they would certainly look to this Article

One of the first paragraphs in the Article, 810.3, states: ?Coaxial cables that connect antennas to equipment shall comply with Article 820?. Since the Wildblue system uses coax, I read this to mean that the sections that cover antenna lead-in conductors in 810 do not apply, but instead 820 is used. For instance "antenna discharge units" are not required. Instead, per Article 820, the requirement is that the shield be grounded. The coax ground block grounding conductor must be insulated per Article 820, unlike Article 810 which allows it to be uninsulated (why are they different?). Also, per part Article 820, an additional ground rod must be installed if the grounding conductor would otherwise be longer than 20 ft., and this new rod must be bonded to existing GES by a #6 wire.

Also According to 810.3, 820.103 is applicable: ?Unpowered equipment and enclosures or equipment powered by the coaxial cable shall be considered grounded where connected to the metallic cable shield?. According to this, then, is not the TRIA grounded through the coax shield, and a separate grounding conductor is not required for it?

Also, there is 810-21(I), Common Ground, ?A single grounding conductor shall be permitted for both protective and operating purposes?. The coax shield serves as an operating (e.g., signal) ground. According to this paragraph, it may also serve as the protective ground required by this section. This clause just recognizes the physical reality that even if another grounding conductor is installed, much if not most of the surge or fault current would flow over the coax shield.

Now some have pointed to the conductor size requirements in 810.21 (H): ?The grounding conductor shall not be smaller than 10 AWG copper, 8 AWG aluminum, or 17 AWG copper-clad steel or bronze.? The shield of RG-6 coax has a resistance of about #12 to #14 copper, depending on vendor. Since this is greater resistance than #10, some installers are arguing that this in not sufficient, despite the language in 820.103. However, the coax shield has less resistance than #17 CCS, which is allowed under 810.21(H). I believe this is due to the fact that the size requirement in this paragraph is driven primarily by mechanical strength rather than electrical resistance. A piece of RG-6 should meet the mechanical strength requirement, and the electrical resistance requirement, of this paragraph.

Finally, there is the installation manual?s instructions to bond dish and TRIA together with a #10 Cu wire. I suppose that, to the extent that components of the antenna can be considered "Unpowered equipment and enclosures or equipment powered by the coaxial cable", they can, therefore be considered grounded. One should then be able to argue that a grounding conductor is not needed for equipment that is already considered grounded?. It?s not even clear to me that this bond wire is needed, as the TRIA, boom, dish and mast are all metal, and are effectively bonded together when assembled.
 

eprice

Senior Member
Location
Utah
Re: Grounding satellite systems thru coax shield

Originally posted by BarryO:
One of the first paragraphs in the Article, 810.3, states: ?Coaxial cables that connect antennas to equipment shall comply with Article 820?. Since the Wildblue system uses coax, I read this to mean that the sections that cover antenna lead-in conductors in 810 do not apply, but instead 820 is used.
But the mast still needs to be grounded in accordance with 810.15.


Also According to 810.3, 820.103 is applicable: ?Unpowered equipment and enclosures or equipment powered by the coaxial cable shall be considered grounded where connected to the metallic cable shield?.
This would not override the requirement in 810.15 that the grounding of the mast comply with 810.21

Now some have pointed to the conductor size requirements in 810.21 (H): ?The grounding conductor shall not be smaller than 10 AWG copper, 8 AWG aluminum, or 17 AWG copper-clad steel or bronze.? The shield of RG-6 coax has a resistance of about #12 to #14 copper, depending on vendor. Since this is greater resistance than #10, some installers are arguing that this in not sufficient, despite the language in 820.103. However, the coax shield has less resistance than #17 CCS, which is allowed under 810.21(H). I believe this is due to the fact that the size requirement in this paragraph is driven primarily by mechanical strength rather than electrical resistance. A piece of RG-6 should meet the mechanical strength requirement, and the electrical resistance requirement, of this paragraph.
For me, this is a key issue. You make a good argument that the coax with it's shield is equivalent in conductivity and strength to a conductor that would comply with 810.21(H). I am nearly convinced.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
Re: Grounding satellite systems thru coax shield

I believe that you are wrong in your assumption about Article 810.

810.1 clearly covers "dish" attenae.
820.1 clearly covers coxial distribution cales of community (CATV) systems only.

So the dish and it's support mast must meet the requirements of 810. The coaxial lead-in must comply with 820. The articles are not mutually exlusive.
 

gndrod

Senior Member
Location
Ca and Wa
Re: Grounding satellite systems thru coax shield

BarryO,
Is the modem ground 'bonded' to the service disconnect main ground bus? Are the two coax lines shields bonded to the modem and TRIA? Are the TRIA and mast bonded together with #10 AWG equivalent surface bonding? If so, there may be a good case for compliance to NEC requirements.
rbj, Seattle
 

gndrod

Senior Member
Location
Ca and Wa
Re: Grounding satellite systems thru coax shield

Is there a reason why masts are not available in composite material. This could solve the grounding issue in this particular instance, wouldn't it?
 

BarryO

Senior Member
Location
Bend, OR
Occupation
Electrical engineer (retired)
Re: Grounding satellite systems thru coax shield

Originally posted by gndrod:
BarryO,
Is the modem ground 'bonded' to the service disconnect main ground bus?
The installation manual specifies that the coax shields be connected to a #14 Cu insulated conductor at the point of entry to the bulding premises, said conductor tied to the electric service GES and no more than 20' in length. If this is not possible, than it is tied to a new ground electrode, which is then bonded to the GES with a #6 Cu wire. In other words, all according to Article 820.
Are the two coax lines shields bonded to the modem and TRIA?
By design, yes. The coax connectors' outer shell threads onto the connectors on the TRIA, which are bonded to the TRIA metal case.
Are the TRIA and mast bonded together with #10 AWG equivalent surface bonding?
The installation manual specifies that a #10 Cu wire be run from a GND screw on the TRIA case, to a bolt on the dish mount.
If so, there may be a good case for compliance to NEC requirements.
rbj, Seattle
'seems so to me, even with the understanding that the 810-15 requirement for mast "grounding" still applies. It's just that most haven't seen the approach of using the coax shields as the conductor specified in 810-21 before. Given that 810-21(i) says that a signal ground conductor (such as the shields) can be used for this purpose, it should be OK, I would think.
 
Re: Grounding satellite systems thru coax shield

Unfortunately the NEC 70 and NFPA 780, even when used together as intended, provide only the minimum protection from lightning damage.

Users of sensitive electronic equipment such as radios, digital tv, stereos, etc that are connected to external equipment on the roof should take heed. Just because compliance with the code 'may' prevent your house from burning down, it may also offer little or no protection from equipment damage.

Such is the case with the highly unadvisable practice of using coax shield as a safe director for energy from even a nearby lightning strike. If this somehow complies with NEC which I hope it does not, it is just another loophole that could result in injury or death if it was exploited.

All masts require a separate and suitable downconductor of the shortest and least geometry possible to a 10' ground rod. The coaxial shield must bond to the downconductor on the ground level at a minimum. That ground rod should bond with no less than #6 copper to the service entrance ground rod. To do any less, regardless if code allows it, is inviting disaster to equipment and/or personnel.

No coaxial lightning surge protection device is intended to handle more energy than the coax center conductor + shield is capable of carrying. For most coax, this is about 5500v under normal conditions. However, coax surge arrestors are very unreliable anywhere near this level of voltage. Accordingly, the coax shield must be grounded at least once before the arrestor.

Protection is provided only for the center conductor, which as the arrestor "fires", it bonds the center conductor to the coax shield and its immediately available (second) ground connection. It does share some of this extremely high voltage with all bonded equipment cases to which it is later connected. Equipment which is not properly bonded to the single point ground is subject to flashover, defeating the purpose of the surge arrestor.

Tens of thousands of insurance and equipment warranty claims are filed each year where applications of grounding, bonding and surge arrestors was improper. And I will suggest that any application relying on coax shield alone (bonded only to the indoor electrical wiring), is doomed to failure, or worse.

Jack Painter
Communications Officer
Fifth Coast Guard District
United States Coast Guard Auxiliary
 

BarryO

Senior Member
Location
Bend, OR
Occupation
Electrical engineer (retired)
Re: Grounding satellite systems thru coax shield

Originally posted by oceanaradio:
Unfortunately the NEC 70 and NFPA 780, even when used together as intended, provide only the minimum protection from lightning damage.

Users of sensitive electronic equipment such as radios, digital tv, stereos, etc that are connected to external equipment on the roof should take heed. Just because compliance with the code 'may' prevent your house from burning down, it may also offer little or no protection from equipment damage.
Agreed. Certainly in part of the country prone to lightning, additional measures should be taken. MIL-HDBK-419A provides good information.

Such is the case with the highly unadvisable practice of using coax shield as a safe director for energy from even a nearby lightning strike. If this somehow complies with NEC which I hope it does not, it is just another loophole that could result in injury or death if it was exploited.
Given the sizes of the mast grounding conductor specified by the NEC, and its allowable configuration, much of the current will unavoidably flow over the shields.

All masts require a separate and suitable downconductor of the shortest and least geometry possible to a 10' ground rod. The coaxial shield must bond to the downconductor on the ground level at a minimum. That ground rod should bond with no less than #6 copper to the service entrance ground rod. To do any less, regardless if code allows it, is inviting disaster to equipment and/or personnel.
These requirements here do go above and beyond those in Article 810, which just specifies a mast grounding conductor as short as practical, but with an unspecified maximum length.

No coaxial lightning surge protection device is intended to handle more energy than the coax center conductor + shield is capable of carrying. For most coax, this is about 5500v under normal conditions. However, coax surge arrestors are very unreliable anywhere near this level of voltage. Accordingly, the coax shield must be grounded at least once before the arrestor.

Protection is provided only for the center conductor, which as the arrestor "fires", it bonds the center conductor to the coax shield and its immediately available (second) ground connection. It does share some of this extremely high voltage with all bonded equipment cases to which it is later connected. Equipment which is not properly bonded to the single point ground is subject to flashover, defeating the purpose of the surge arrestor.
Note that there is no requirement in the NEC for coax surge arrestors, only that the shield be grounded to the GES at the point of entry to the structure.
 
Status
Not open for further replies.
Top