adding surge protectors in LB connector

[tmbrown97]

FWIW, I've done some creative connections just like you describe with LB boxes...  I've tapped the side to do a water-tight cable fitting...  with larger ones I've punched a larger hole in the side and done the threaded conduit fitting with the nut on the inside.  It's totally doable, but I'm not sure how widely acceptable it would be to the electrical contractors.  Luckily I only deal with LV and the only reason I'm dealing with the stuff is for structure protection (my wire is rated for sun exposure, underground, etc).
 
All that said, if you have the option of just doing a box inside, that's much cleaner.  I have to do the Cat5 surge protection pretty often on my outdoor microwave installations.  For the stuff I do, it's critical that you use shielded CatX wire and properly tie in the shielding and the copper drain wire into the surge device (I use polyphaser generally because it's spec'd by the manufacturers).  I then try to get electrical ground as quick as I can and mount the polyphaser either right at the building entrance or as close to it as I can possibly get.

 

[mikefamig]

As fate would have it thunderstorms came through my back yard last night waking me up to a huge bang. The power went off for about an hour and when it came back on the wall wart for a netgear switch was dead. Everything else seems ok so far but I won't know if the switch was damaged until I get power to it to test it.
 
I better get on the stick with installing protection for the Elk!
 
Mike.

 

[mikefamig]

After doing some more research I decided to put the surge protector in the can with the elk control grounding it to the AC outlet that is installed there. While it is about 25' from the house's' load center and ground rod the AC outlet has a direct path back to the load center on it's own dedicated circuit.
 
I learned that a general rule for surge protection is to have a shorter path to a good ground than to the device that it is protecting. The question becomes what exactly is considered a good ground. I also learned that a "good ground" is defined as having less than 25 ohms resistance to earth and preferably less than 5 ohms. I feel that a 25' long awg 14 romex fits that description having less than even 1 ohm of resistance.
 
So to sum it up I just put the Ditek in the same enclosure with the M1 control using a 6' wire between the Ditek and the control to give the Ditek time to switch.
 
Does anyone see a problem with this> Is my logic good?
 
Mike.

 

[LarrylLix]

Just an observation.
 
Lightning is composed of high frequency bursts. People like to have extra conductor coiled up to take up it's slack and not shorten to fit. This air coil can act as a choke to high frequencies.
 
In a bad lightning hit the high voltage will just jump across the coil but in lower voltage lightning inductions the coil will just impede the discharge to ground unnecessarily raising the differential voltage between your equipment, ground, and other conductors.
 
Passing lightning discharge conductors through ferrous holes can be an even larger impedance, acting as a high-frequency choke.
 
 
If you installed this in my area, with winter frost, the ground would have fractured all your conduits or broken the LBs out of the wall as the ground pulls down on the pipe. Sand padding and/or sleeves over the pipe entering the ground would help. An elbow under the ground would make the problem worse giving dirt something to pull on.

 

[mikefamig]

Just an observation.
 
 
If you installed this in my area, with winter frost, the ground would have fractured all your conduits or broken the LBs out of the wall as the ground pulls down on the pipe. Sand padding and/or sleeves over the pipe entering the ground would help. An elbow under the ground would make the problem worse giving dirt something to pull on.

 
The conduit was installed by a "professional contractor". How would you avoid this frost heave problem moving the conduit? Maybe a flexible pvc elbow in place of the l-box?
 
Mike.

 

[LarrylLix]

http://www.homedepot.ca/webapp/wcs/stores/servlet/SearchView?catalogId=10051&storeId=10051&langId=-15&N=0&Ntt=expansion+PVC&Nty=1&D=expansion+PVC&Ntx=mode+matchallpartial&Dx=mode+matchallpartial&s=true

There are expansion sleeves available but they are expensive and not easy to retrofit.
 
If frost is a problem and these conduits are just stubs entering the ground without an elbow at the bottom. I would dig around them and attempt to slide a larger split piece of PVC around them. With elbows under the ground we usually pad with granular material like sand or gravel sand mix so that with any sinkage the material can flow around the pipe and not pull.
 
My neighbour (and  electrician too) did this through his sidewalk slab and his pipe is broken exposing his conductors. The sidewalk slab on his new house sank several times and made a 2" gap where the conduit break is. Of course concrete has no give at all. (duh!).
 
The contractors would always rigidly fasten meter bases on the side of people's houses to attempt to prevent the pull down. This resulted in the brick being broken and the base still floating in the air. Now, with many complaints, the base manufacturers started to put a loop space at the top of the base where electricians can make crossed loops of slack, if they know what the space is for. The old conductor loop in the ground trick only made it worse as it gave the ground something to pull on.
 
Is this a recent installation that you haven't gone through a few winters yet?

 

[mikefamig]

http://www.homedepot.ca/webapp/wcs/stores/servlet/SearchView?catalogId=10051&storeId=10051&langId=-15&N=0&Ntt=expansion+PVC&Nty=1&D=expansion+PVC&Ntx=mode+matchallpartial&Dx=mode+matchallpartial&s=true

There are expansion sleeves available but they are expensive and not easy to retrofit.
 
If frost is a problem and these conduits are just stubs entering the ground without an elbow at the bottom. I would dig around them and attempt to slide a larger split piece of PVC around them. With elbows under the ground we usually pad with granular material like sand or gravel sand mix so that with any sinkage the material can flow around the pipe and not pull.
 
My neighbour (and  electrician too) did this through his sidewalk slab and his pipe is broken exposing his conductors. The sidewalk slab on his new house sank several times and made a 2" gap where the conduit break is. Of course concrete has no give at all. (duh!).
 
The contractors would always rigidly fasten meter bases on the side of people's houses to attempt to prevent the pull down. This resulted in the brick being broken and the base still floating in the air. Now, with many complaints, the base manufacturers started to put a loop space at the top of the base where electricians can make crossed loops of slack, if they know what the space is for. The old conductor loop in the ground trick only made it worse as it gave the ground something to pull on.
 
Is this a recent installation that you haven't gone through a few winters yet?

 
This is solid conduit with an elbow below the surface and it has been installed since 2011 with no problem yet. If it does become a problem I suppose that I will dig around it and fill with some loose material.
 
Mike.
 
EDIT
 
Correction.....the elbow below the surface is actually a 90 degree sweep.

 

[Neurorad]

I'm confused about grounding and surge protection.
 
All cables entering a house need to be grounded.  I think this is different from surge protection.  Does that Ditek unit provide a direct grounding connection, as well as surge protection?
 
As an aside, Is the outbuilding bonded to the house?  That is, will the ground rod at the outbuilding be at the same potential as the house?
 
I've been learning about grounding and bonding piecemeal, over the last several years.  I understand the concepts, but I'm unfamiliar with the details of how they are accomplished.
 
Based on my reading, I think the ideal method to minimize surges from an incoming signal cable would be to isolate the circuits optically - similar to using fiberoptic cable for Ethernet.  I'm unsure why this isn't a realistic option. 
 
Perhaps it has to do with the inability  to optically isolate power?  Signalling usually uses a reference to ground, AFAIK.

 

[LarrylLix]

Grounding is quite the art.
 
Signalling really wants no grounding on inputs at all. Floating differential inputs are the best way to avoid noise as differences in conductors run in parallel physically have common mode disturbances on them and those inputs tend to eliminate them. This is good for lower voltage problems which, hopefully by the time they hit your inputs are at low levels.
 
For grounding and bonding think about the eqi-potential thing. People always think of reducing unwanted surges and spikes to ground level. Think of bring the ground up to the noise level or more appropriately, keeping the ground and the conductor at the same potential. When you tie the two together it doesn't matter what voltage it is at, there is no difference to bother sensitive circuits.
 
The problem comes into your "floating voltage island" when you run signal out of your zone. Telephone wiring, cable TV, Ethernet. antenna, or whatever bring somebody else's "floating voltage island" into your electronics and now the differences are damaging. There is no such thing as "ground", only YOUR ground.
 
When your utility high voltage lines have a fault to ground the ground at that point is at the potential somewhere less than the phase voltage. The voltage back at the feeding transformer's ground may be considered zero. If you are somewhere in between you have a ground potential rise above the transformer's ground level. A phone companies cable coming from that area brings that in and you have the difference across your equipment. You may even experience some difference across your property killing you cows or (I know of an actual case from a Lineman) tingling your balls.

 

[Del91574]

After doing some more research I decided to put the surge protector in the can with the elk control grounding it to the AC outlet that is installed there. While it is about 25' from the house's' load center and ground rod the AC outlet has a direct path back to the load center on it's own dedicated circuit.
 
I learned that a general rule for surge protection is to have a shorter path to a good ground than to the device that it is protecting. The question becomes what exactly is considered a good ground. I also learned that a "good ground" is defined as having less than 25 ohms resistance to earth and preferably less than 5 ohms. I feel that a 25' long awg 14 romex fits that description having less than even 1 ohm of resistance.
 
So to sum it up I just put the Ditek in the same enclosure with the M1 control using a 6' wire between the Ditek and the control to give the Ditek time to switch.
 
Does anyone see a problem with this> Is my logic good?
 
Mike.

Bad idea.
 
I'll put it as this Mike, we're dealing with a large scale lightning suppression installation literally around the corner from you (large scale Biopharma site, 1 Million square feet for the main building alone) and the conductors that are being installed are bonded straight to building steel which is directly connected to the grounding grid for the site, connected straight to the building ground rods themselves. I'm battling the guys that are installing air terminals literally on top of my PTZ cameras/goosenecks on the roof (think about what cabling is involved with an analog PTZ before it hits a FOC).
 
Difference between bonding and grounding is what is key here. Surge should be connected using heavy enough cable directly to the building EG (ground rod, ring, grid, etc.). The ground installed in a NM cable is for overcurrent fault protection and not surge. While it's better than nothing, if the only purpose is to save a length of cable of 25-30', it's not the best idea.
 
In your case/installation, the best scenario would be to install a pair on each end of the LV cabling, at the panel side and the remote side.

 

[mikefamig]

Bad idea.
 
I'll put it as this Mike, we're dealing with a large scale lightning suppression installation literally around the corner from you (large scale Biopharma site, 1 Million square feet for the main building alone) and the conductors that are being installed are bonded straight to building steel which is directly connected to the grounding grid for the site, connected straight to the building ground rods themselves. I'm battling the guys that are installing air terminals literally on top of my PTZ cameras/goosenecks on the roof (think about what cabling is involved with an analog PTZ before it hits a FOC).
 
Difference between bonding and grounding is what is key here. Surge should be connected using heavy enough cable directly to the building EG (ground rod, ring, grid, etc.). The ground installed in a NM cable is for overcurrent fault protection and not surge. While it's better than nothing, if the only purpose is to save a length of cable of 25-30', it's not the best idea.
 
In your case/installation, the best scenario would be to install a pair on each end of the LV cabling, at the panel side and the remote side.

 
Thanks for the input an dmore questions....
 
1 - On the remote/garage side I have the Elk enclosure adjacent to the load center and have grounded the Ditek to the load center's ground which is connected to a stake in the earth which I think that you are saying is sufficient protection, no?
 
2 - On the house side I do have a steel I-beam construction and can run a awg 6 wire from the control enclosure to the I-beam in the basement below it without much difficulty. Would you consider that a good ground for surge protectoion?

 

[Del91574]

Connect directly to the ground rod where possible. If not, as close to it on the main feeder from it inside. Bypass the load center.
 
Doubtful your building steel is connected to EG.
 
If you don't have a choice, it's better than nothing, but if it's accessible, bond it right.

 

[mikefamig]

Connect directly to the ground rod where possible. If not, as close to it on the main feeder from it inside. Bypass the load center.
 
Doubtful your building steel is connected to EG.
 
If you don't have a choice, it's better than nothing, but if it's accessible, bond it rig

 
You say connect directly to the ground rod but what exactly is it that is connected to the ground rod. There has to be some length of conductor between the Ditek and the ground rod but what gauge of conductor is sufficient to connect the Ditek to the ground rod and at what length? If 25' of awg 14 as I have it now is not direct enough then what gauge of wire is good enough over 25'?

 

[Del91574]

The Ditek, via 12AWG would be my suggestion. How you get it to the rod and protect it from physical damage is another consideration. Could do compression to the grounding conductor, but definitely bypass the load center portion.
 
Best case scenario is to the ground rod and to the same connection, cad welded or similar, but usually neighbors don't like it when you do that.

 

[Neurorad]

How about another ground rod, at this cable entrance, bonded to the existing ground rod at the service entrance? Diteks grounded to the 2nd ground rod.