Grounding for -Safety- and -Lightning-

[flippinfleck]

Keep in mind that I am not doing any grounding for RF at this point.

Looking to put three masts up on the roof (RX, TX, and future RX/TX) and want to make sure that, from a safety perspective, I'm on the right track. Talked to a couple of electricians and they were all about not connecting antenna ground to service ground. Everything I can read on the interwebs says that they should be connected...

Any thoughts on this preliminary plan? It may be overkill but that's what I'm looking for.

 

[cockpitbob]

I'm assuming you'll also ground the rig and power supply to your ground system. Since your power supply almost certainly has a 3 pin plug, you are connecting to the AC service ground no matter what. I would not add additional grounding to the service ground beyond that.

If I understand your diagram, you'll have one of the best grounded set of antennas anywhere. 7 ground rods is a lot and probably overkill. Make sure you run the cables from the antennas to rods outside the house. You do not want lightning coming inside. (probably an obvious statement)

One thing to bear in mind is that nothing can prevent radio damage from a direct lightning strike. There's way too much energy to deal with. Those ground cables and rods will prevent damage to your house. Those little coax arrestors we use are to save the radio from the EMP of a close strike. If there's a direct strike they'll be a molten blob inside. A ham here (Zappa I think) tells a story of disconnecting the coax from his rig and laying it on a steam radiator. When lightning struck nearby, a big spark would jump from the connector to the radiator(through the radiator's paint). There's a lot of RF energy radiated in a strike. I always disconnect the coax when I think there will be lightning. Sometimes thunder wakes me up and if I haven't disconnected the cables I have to decide if I want to take the risk of having my hands on the cable when the antenna gets hit

 

[n1bsbri]

Keep in mind that I am not doing any grounding for RF at this point.

Looking to put three masts up on the roof (RX, TX, and future RX/TX) and want to make sure that, from a safety perspective, I'm on the right track. Talked to a couple of electricians and they were all about not connecting antenna ground to service ground. Everything I can read on the interwebs says that they should be connected...

Any thoughts on this preliminary plan? It may be overkill but that's what I'm looking for.

I don't know why electricians who should know better say stuff like that. Many references (including the National Electrical Code, IEEE Standards, Motorola R-56 install standards, etc) state that ALL grounds must be bonded together, electrical, water, phone, CATV, etc.).

Here's one layman's reference that states it pretty clearly, there are many others out there that can be found with a quick Google search.

page 8 of this document, especially the 4th paragraph on. Actually the whole document is worth a read.

Ham Radio Station Protection

 

[AHM]

... I always disconnect the coax when I think there will be lightning. Sometimes thunder wakes me up and if I haven't disconnected the cables I have to decide if I want to take the risk of having my hands on the cable when the antenna gets hit

Spend .59¢ and leave the lead-in disconnected except when you are using it.

73's,

 

[Chet0729]

Multiple ground points increase your exposure to lightening strike damage and also introduce other safety issues . Single point grounding is the best solution to lightening protection. If something happens on the ground you want it to happen on all the grounds or you equipment will become the fuse and that is never good.
i took a lot of classes related to this when I was doing POS installs and later in some computer room design seminars.
You will never stop lightening but a properly designed grounding system can minimize damage to equipment.

 

[drgrant]

Spend .59¢ and leave the lead-in disconnected except when you are using it.

73's,

If I found any of those in my junk boxes they would get thrown away. 83-1SPs only please.

-Mike

 

[Mike-Mike]

Spend .59¢ and leave the lead-in disconnected except when you are using it.

73's,

I always got a kick out of this... The lighting travels a half mile through the atmosphere and wont jump 3" from the disconnected coax to the grounded radio?

 

[cockpitbob]

I always got a kick out of this... The lighting travels a half mile through the atmosphere and wont jump 3" from the disconnected coax to the grounded radio?

More than 3" in my case, but for a direct hit you're right, a whole lot of stuff in the house is toast. It's the near misses that these measures help with.

 

[Chet0729]

More than 3" in my case, but for a direct hit you're right, a whole lot of stuff in the house is toast. It's the near misses that these measures help with.

Exactly. it is the close by hits that blow stuff up that is not properly grounded. This has happened twice at least at the club i shoot at because the installation company that put in the electronic gate did it wrong.

 

[Rob Boudrie]

Exactly. it is the close by hits that blow stuff up that is not properly grounded. This has happened twice at least at the club i shoot at because the installation company that put in the electronic gate did it wrong.

Do you shoot at Hopkinton?

We lost both card system controllers (each controls 4 doors); 2 readers on one unit and one reader on the other unit in a recent lightening storm. There were no signs of lightening strikes, but one of out photoeye operated entrance lights also got whatcked, but our computer was fine. We had a surge suppressor on the access systems, bur have since added a UPS and will add a whole panel surge protector.

I was told to get a lightening proof system instead, however, the person who suggested this was unable to offer a specific brand/model and tole me he was leaving the details to me

 

[AHM]

Two from the vaults

If I found any of those in my junk boxes they would get thrown away. 83-1SPs only please.

Consider the human factors. Make it hard enough to disconnect the lead-in that the operator blows it off, and after the first brush with Zeus, maybe they can throw the entire shack (or house) away.

---

I always got a kick out of this... The lighting travels a half mile through the atmosphere and wont jump 3" from the disconnected coax to the grounded radio?

The lightning travels a half mile through the atmosphere and would rather jump from the coax center conductor 3" to the rig than ½" to the grounded shield? Maybe the nuance is that it built up momentum on the way down from the cloud, hi hi?

But I've given you only half a loaf with that coax disconnect. The missing link is a vital piece of gear: the mayonnaise jar.

73's

 

[flippinfleck]

I'm assuming you'll also ground the rig and power supply to your ground system. Since your power supply almost certainly has a 3 pin plug, you are connecting to the AC service ground no matter what. I would not add additional grounding to the service ground beyond that.

Currently, the only service ground is a 4Ga cable from the inside panel to the water main. The plan is to tie in the additional "ground groups" to the existing ground system. In the front (left on the diagram) by directly attaching to the 4Ga wire. On the side (top on the diagram), by bonding to the outside faucet. I've already confirmed that it's copper all the way to the water main.

If I understand your diagram, you'll have one of the best grounded set of antennas anywhere. 7 ground rods is a lot and probably overkill. Make sure you run the cables from the antennas to rods outside the house. You do not want lightning coming inside. (probably an obvious statement)

The only ground that would be internal is the connection on the front side. Since there is no external ground currently connected to the electrical mains, it has to make a short entrance through the exterior wall. I chose this spot for it's proximity to the existing mains ground. It was either that or run it all the way around the house to the external faucet. My wife will not like me digging up the driveway.

Would that short internal run pose a greater risk than the probably 75' run of grounding to get from the front to the side?

One thing to bear in mind is that nothing can prevent radio damage from a direct lightning strike.

I am only looking to help with non-direct strikes. While my roof is not the tallest thing in the area even with a 10' mast, it's within 20'. I'm hoping that all the trees (pine and oak) within a block of me take the brunt and a overbuilt ground system on my property prevents my place from burning to the ground.

Multiple ground points increase your exposure to lightening strike damage and also introduce other safety issues . Single point grounding is the best solution to lightening protection. If something happens on the ground you want it to happen on all the grounds or you equipment will become the fuse and that is never good.
i took a lot of classes related to this when I was doing POS installs and later in some computer room design seminars.
You will never stop lightening but a properly designed grounding system can minimize damage to equipment.

Understood about a direct strike and the need to bod all grounding systems. It was confusing to hear electricians (both union and non-union) tell me otherwise.

More than 3" in my case, but for a direct hit you're right, a whole lot of stuff in the house is toast. It's the near misses that these measures help with.

When I was probably about 7 or 8, our house took a near-direct hit. The lightning struck the church steeple across the street and arced from there to our house (as well as a very large pine tree almost a block away). We found bits of chimney bricks spread out at least 100' in all directions and lost several appliances. The pine tree was split down the middle. Not hoping to stop direct hits, just mitigate issues if something close by gets hit.

 

[Chet0729]

Do you shoot at Hopkinton?

We lost both card system controllers (each controls 4 doors); 2 readers on one unit and one reader on the other unit in a recent lightening storm. There were no signs of lightening strikes, but one of out photoeye operated entrance lights also got whatcked, but our computer was fine. We had a surge suppressor on the access systems, bur have since added a UPS and will add a whole panel surge protector.

I was told to get a lightening proof system instead, however, the person who suggested this was unable to offer a specific brand/model and tole me he was leaving the details to me

No WPRC but I'm sure you have the same issues. Data lines connecting the gate equipment to the main building through an Ethernet cable. At our club they made it even worse by installing a separate electrical service for the gate. Lazy electricians, there is no other excuse. Most of these electricians know nothing about what needs to be done to protect these systems. I went off at a meeting because the club had to pay over 3K twice because the company that installed the stuff are a bunch of bozo's. They recommended very expensive surge suppressors which will not solve the issue. After I explained everything to the president and he questioned them they proposed the same solution I had. They didn't feel bad about charging us and until I proposed a fix they had no options.As far as I know they have not fixed it yet. I was so pissed when I left the meeting. These idiots put in pipe for data and power and didnt put in any additional pull wires. This is electrician 101, you lways put in pull wires.

 

[Chet0729]

Do you shoot at Hopkinton?

There were no signs of lightening strikes, but one of out photoeye operated entrance lights also got whatcked, but our computer was fine. We had a surge suppressor on the access systems, bur have since added a UPS and will add a whole panel surge protector.

I was told to get a lightening proof system instead, however, the person who suggested this was unable to offer a specific brand/model and tole me he was leaving the details to me

Surge suppressors and UPS's will not solve your issue. There is no such thing as lightening proof but there are a lot of basics that can make it a lot less susceptible. if you would like to discuss it some time let me know.

 

[flippinfleck]

This is electrician 101, you lways put in pull wires.

When I ran data and coax for the house I made sure there were pulls left in the walls. You don't need to be an electrician to know that doing the same job twice is a waste of time.

 

[Chet0729]

Currently, the only service ground is a 4Ga cable from the inside panel to the water main. The plan is to tie in the additional "ground groups" to the existing ground system. In the front (left on the diagram) by directly attaching to the 4Ga wire. On the side (top on the diagram), by bonding to the outside faucet. I've already confirmed that it's copper all the way to the water main.


The only ground that would be internal is the connection on the front side. Since there is no external ground currently connected to the electrical mains, it has to make a short entrance through the exterior wall. I chose this spot for it's proximity to the existing mains ground. It was either that or run it all the way around the house to the external faucet. My wife will not like me digging up the driveway.

Would that short internal run pose a greater risk than the probably 75' run of grounding to get from the front to the side?


I am only looking to help with non-direct strikes. While my roof is not the tallest thing in the area even with a 10' mast, it's within 20'. I'm hoping that all the trees (pine and oak) within a block of me take the brunt and a overbuilt ground system on my property prevents my place from burning to the ground.


Understood about a direct strike and the need to bod all grounding systems. It was confusing to hear electricians (both union and non-union) tell me otherwise.


When I was probably about 7 or 8, our house took a near-direct hit. The lightning struck the church steeple across the street and arced from there to our house (as well as a very large pine tree almost a block away). We found bits of chimney bricks spread out at least 100' in all directions and lost several appliances. The pine tree was split down the middle. Not hoping to stop direct hits, just mitigate issues if something close by gets hit.

Non direct strikes with the grounding you show in you diagram will almost certainly lead to blown up equipment.

 

[flippinfleck]

Non direct strikes with the grounding you show in you diagram will almost certainly lead to blown up equipment.

Care to elaborate? I'm not as concerned about the equipment as I am about fire.

 

[n1bsbri]

Care to elaborate? I'm not as concerned about the equipment as I am about fire.

Your ground rods need to be bonded together and to the utility ground system. It's the voltage gradient developed across these spacially separated ground systems that blows up equipment.

 

[Chet0729]

Care to elaborate? I'm not as concerned about the equipment as I am about fire.

Basically when there is a strike that is close by and closer to one of your ground points than the other there can be a huge difference in potential between your ground points for a very short time. This difference in potential on the grounds, if the equipment is tied together electrically somehow, will burn equipment up. This can happen when you have something like coax, ethernet, phone lines etc connecting two potential ground connections. There are two ways to cure this issue, either common point grounding needs to be used or a non metalic type of connection needs to be used. The non metalic connection is easy in the communication world by using fiber optics, it is not as easy in the radio world. I would strongly suggest you run a ground from your radio tower back to the ground point for your home electrical panel.

 

[flippinfleck]

Your ground rods need to be bonded together and to the utility ground system. It's the voltage gradient developed across these spacially separated ground systems that blows up equipment.

The intent is to have all systems bonded. All ground rods on the left of the diagram would be bonded together, all ground rods in the top of the diagram would be bonded together, and then each group would be bonded to mains. Would I need to bond both ground systems to each other in addition to bonding to utility (mains)?

 

[Chet0729]

Your ground rods need to be bonded together and to the utility ground system. It's the voltage gradient developed across these spacially separated ground systems that blows up equipment.

Even this will cause potential issues. Anytime more than 1 ground rod is used there is potential for equipment damage.

 

[flippinfleck]

Even this will cause potential issues. Anytime more than 1 ground rod is used there is potential for equipment damage.

The electrical panel is currently connected to the water mains VIA a 4Ga copper stranded wire (depicted by a dashed dark green. currently _inside_ the house). I intend to connect the ground rods on the left of the diagram to each other with 6Ga copper and then directly to this 4Ga copper stranded wire VIA 6Ga copper. Same thing for the ground rods at the top of the diagram, except these will be connected to an external copper pipe with 6Ga copper that runs straight back to the water mains.

Is that not common-point grounding? Everything is tied to the electrical mains.

 

[n1bsbri]

Even this will cause potential issues. Anytime more than 1 ground rod is used there is potential for equipment damage.

NEC 800 requires a separate grounding conductor if the surge protection device is over 20' from the ground point. The key is to bond all of the ground rods together.

 

[n1bsbri]

The intent is to have all systems bonded. All ground rods on the left of the diagram would be bonded together, all ground rods in the top of the diagram would be bonded together, and then each group would be bonded to mains. Would I need to bond both ground systems to each other in addition to bonding to utility (mains)?

Commercial standards would have external ground connections bonding all grounding electrodes together (and bonded to the inside grounding system). It might be overkill for home use, as long as you have all the points connected together.

 

[flippinfleck]

Commercial standards would have external ground connections bonding all grounding electrodes together (and bonded to the inside grounding system). It might be overkill for home use, as long as you have all the points connected together.

I think I follow; the lack of a direct connection (outside of both being connected to the electrical mains ground) between the two groups of ground rods would be an issue causing a difference in ground potential between the two groups.

Am I close?

 

[n1bsbri]

I think I follow; the lack of a direct connection (outside of both being connected to the electrical mains ground) between the two groups of ground rods would be an issue causing a difference in ground potential between the two groups.

Am I close?

Yes, in a surge event, there is a voltage gradient that is developed between ground rods that have distance between them. It can be 100K volts or more. If that 100K volts is allowed to develop between the shield of the antenna coax, and the electrical safety ground of the equipment, smoke comes out of your radio. The idea is to provide a bypass route that "shorts out" that gradient and keeps it to a level that won't damage your equipment. It may not protect against a direct strike, but most surge events are "near misses" anyway.

 

[Rob Boudrie]

I went off at a meeting because the club had to pay over 3K twice because the company that installed the stuff are a bunch of bozo's. They recommended very expensive surge suppressors which will not solve the issue. After I explained everything to the president and he questioned them they proposed the same solution I had.

I inherited care of the system from the former club pres (licensed electrician) who is now deceased. I replaced the system (an ancient one that still used Weigand swipe cards) and used his wire to the readers that are laid in conduit - I don't think there are pull wires, but I know he used fish tape rather than pull wires for the original pull.'

I guess we got off easy - each motherboard is a flat rate $440 to exchange for a refurb board, plus fedex fees. I suppose it probably would have been a few thousand if we were using a company.

Can you describe your solution in detail? I am very interested.

 

[flippinfleck]

Forgive me if I sound pedantic, or if I'm just plain dense. Two questions:

If I added a direct connection between the two groups of ground rods (even if this direct connection consisted of an additional four ground rods inline (12'-16' between ground rods) as well as the copper cabling required), this configuration would no longer pose an issue?

If I removed one of the groups of ground rods and instead used better feed line to make up for longer coax runs, would this be better? The current runs are about 75' end to end (grounding block near the building entrance for coax) for receive only.

 

[flippinfleck]

Can you describe your solution in detail? I am very interested.

As would I. I would personally have used a wireless bridge if the gate had it's own service...

 

[Rob Boudrie]

As would I. I would personally have used a wireless bridge if the gate had it's own service...

Not sure that would have saved the systems I had to replace. I don't even have an indication that the lightening hit the gates, and think it may have been a megasurge that the consuler grade surge protector could not handle. I heard that a commercial facility just down the street lost two furnaces and an A/C unit the same night.

One had no connection to the outside gates, and the only common connection between the two was the power source and the fact that both were plugged into the same router - and curiously not only did the router survive, but the IP communications with the control boards was fine - the reader ports were non-functional, and 1 of the 2 readers that was trashed was not on a gate.

It is interesting that you mention wireless .... google tells me that Snap Wireless makes a RS422 unit that I could use at $104 per endpoint. The locks still use power switched through the system, but those are on BFIRs (large isolation relays triggerd by the smaller relay on the controllers).