I recently had a lighting strike and lost about $1000 worth of equipment. I’d like to reduce the chance of that happening again so I’m looking for advice.
I have a UDM in my house, with a 125 foot run underground in conduit to my barn. In the barn, I have a POE switch that feeds 10 cameras and an Ubiquiti AP. I’d like to add a ground somewhere. I just purchased a surge protector with ethernet for the barn, since the switch is currently plugged in directly to an outlet and should be protected anyway. I also bought this from APC for my equipment in the house. I was going to install that between my UDM and POE switch in the house, then ground it to an outlet.
I’m reading so much information about how to go about this. My barn is powered with 220v from my house, so 4 wires go to the barn H/H/N/G. the ground on the barn is the same ground as the house. If I use both devices can that create a ground loop in the event of a surge? I’m also reading that I can use the APC at any point on my network to provide protection. Is this correct?
Please don’t suggest fiber runs, as the cable is already run and I don’t plan on redoing it. Thank you all in advance.
Boxes powered by an electrical source means protection is gone - compromised.
All over the world even over 100 years ago, copper wires were connected to effective (earthed) protectors so that even direct lightning strikes caused no damage. So effective that damage from lightning is routinely considered a human mistake.
If lighting burns out one media converter, that solution is bogus.
Effective protection all over the world (even 100 years ago) meant direct lightning strikes without a surge anywhere inside.
Electronics atop the Empire State Building are struck, typically, 23 times annually. Electronics atop the WTC - about 40 times annually without damage. How can this be when wild speculation says something different?
Or learn from the IEEE. Properly earthed solution is only 99.5% to 99.9% protection. So it is not perfect. Then the IEEE (being an honest and professional organization) says this:
Best protection is routinely on copper wire between buildings. If earth grounds at both ends are properly connected. Either directly or via a protector. Best solution costs tens of times less money.
I love the (close-but-no-cigar) citation of the IEEE document titled “Grounding of Industrial and Commercial Power Systems”, which primarily deals with lightning protection for industrial and commercial buildings, equipment, power substations, and the like.
It’s a good read, if you like air gaps, Faraday cages, and even makes mention of using lasers and rockets to dissipate storm clouds.
Every household should have their own rocket launcher. :-)
No mention of communication cabling whatsoever.
In fairness, section 2.7 does deal with interior wiring systems, but it defers to the NEC with regards to adequate grounding solutions, and, in the first few paragraphs of section 2.7.1, it clearly states:
“Basically, the NEC designates minimum acceptable limits for safety that may not be adequate for a particular application and may not necessarily provide for the efficient or practical use of high technology utilization equipment”
So, the IEEE defers to the NEC standards with regards to proper grounding/protection of interior wiring, and states that the NEC guidelines “might not be adequate for protection of high-technology equipment.”
So, there’s that.
Yeah, OP could put some ethernet arrestors on both ends of the cable, and then of course bond each one to a grounding source where the cables enter into each building. There are plenty advertised online, many of which come with their own sets of negative reviews after the devices failed to do their jobs, with a few of the reviewers being adamant that they followed the proper grounding procedures.
That of course doesn’t mean that some other factors weren’t in play, but it is certainly enough to raise suspicion, while simultaneously raising enough doubt to conclude that the devices might not offer reliable/predictable protection, even when installed correctly.
The IEEE 99.5% rule doesn’t come into play here, as it only covers unrelated grounding scenarios, and doesn’t apply to the real-life performance of the lightning arrestors themselves.
While some arrestors may have failed at their jobs due to incorrect grounding methods, the folks that claimed to have closely followed the included grounding instructions may very well have had a defective unit. Perhaps bad solder joints, a GDT that was compromised due to rough handling during shipping or from too much heat during soldering, or some other internal flaw.
Sadly, there’s not a convenient way for home users to test them, other than the “wait and see” method. Not the best way to find out. :-/
As with any gas-in-a-tube device, leakage can develop from sudden impacts to the chassis, or simply from the passage of time, etc. Once the gas is gone, so is the protection for that conductor, and you only need one to fail for catastrophic results.
I wouldn’t expect to see a manufacturer’s analysis of failure modes and failure rates anytime soon. . . Heheh.
However, you did bring up an interesting point that gave me pause.
“Boxes powered by an electrical source means protection is gone - compromised.”
So you’re positing that lightning could induce a current flow in the underground cable, which will enter into the first media converter via ethernet, through its electronics, and then out to the building’s electrical system via the converter’s AC-adapter, where it will then enter the AC adapter of the second media converter, proceed through its electronics, and out to whatever ethernet device it is connected to.
I can actually envision that path, but can’t even guess at the likelihood of that happening, any more than I could speculate on the failure rate of ethernet lightning arrestors, or the longevity of the gas-filled tubes inside them.
Either event seems possible, and is a matter of percentages.
Thanks for giving me things to ponder. . .
Fiber converters only do something useful when all other incoming wires (especially AC) have well proven protection as implemented over 100 year ago. If every other incoming wire is not properly earthed, then irrelevant is whether a data cable is fiber or copper.
Surges are not induced. Surges are (for example) lightning’s current incoming and outgoing. A direct connection.
Lightning current (a connection from cloud to distant earthborne charges) is even incoming on buried wires. As this Tech Note clearly demonstrates. Even underground wires can be part of lightning’s current path. As demonstrated in the 1950s when telcos did research for COs. That would be filled with germanium transistors.
Surges (including lightning) are always a direct current flowing in and out between two charges. For lightning, a path from charges in a cloud to charges (maybe four miles away) in earth. That path must not connect anywhere inside a structure. For surge protection to exist.
Concepts (also defined elsewhere by the IEEE) dictate a “single point” ground. And when properly earthed, the protection is 99.5% to 99.9%. From all surges - especially direct lightning strikes.
For example, lightning struck a tree. That electric current went 100 feet underground to the building’s earthing electrode. Up into the house. Through appliances. Then out so earth via electrodes on the other side. Then that current flowed some four miles to distant charges.
Damage because a best path to those distant charges was up through appliances in that house. Damage because earthing was not single point.
Only wild speculation claims E-M fields at that tree caused damage 100 feet away. Surges are always a current that directly flows through a structure to cause damage. For lightning, damage is in the path that connects a cloud to distant earthborne charges.
Why were church steeples damaged? Even wood is an electrical conductor. But not a very good one. So wood is damaged. Franklin’s lightning rod (like all surge protection) was only a more conductive path to distant earthborne charges. Protection is always and only about providing a better path, outside, so that surge currents are nowhere inside.
In another venue, they had Fios - fiber optics. Lightning struck cause damage to various networked appliances. Fios ONT also was damaged. How can this be? It is a fiber converter? They did not have properly earthed protection on every incoming wire. So a direct lightning strike out at the pole flowed destructively through the ONT (fiber optic box) and other appliances. To connect to distant charges.
Surge damage is always about the current path into and out of a structure. Protection only exists when that current has a path that need not go into the building.
Box (Fios ONT) “powered by an electrical source means protection is gone - compromised.” Protection only exists when every incoming wire connects to the same earthing electrodes. Either directly or via a protector. Those connection to and earthing electrodes require most all attention. Virtually all professionals say that.
NEC only defines a barest minimal earthing.
Regarded as legendary are Polyphaser’s application notes that discuss this extensively. In one note, lightning struck the manhole cover. Which connected lightning destructively underground into a communication building maybe hundreds feet distant. Not damage induced by induced fields as so many only wildly speculate. Damage is always about where that direct strike current flows. Its incoming and outgoing path. Even on underground cables.