https://blog.jasonantman.com/2018/07/shamelessly-over-engineered-coax-lightning-protector/
Modem, router, some Cat5 runs, and my beloved, never-failed-me Hubitat were wiped out. Modem & router replaced, will tackle Hubitat replacement soon with spare. This is the 2nd cable modem lost to lightning in 3 years, but the first time it has gone through the ethernet. (TP-link network switch still test Ok so far.) I do not have any POE but was looking at something like this Ethernet Surge Protector
https://www.amazon.com/Tupavco-Ethernet-Protector-Gigabit-1000Mbs/dp/B00805VUD8/
I already have a 75 Ohm Lightning Surge Protector installed where the coax enters the distribution cabinet in the garage --obviously not doing a great job.
I asked the EEs & commercial installers for better ideas to up my game re lightning hits as storms are common here. The airgap seemed to be the best and -bonus- the blog above had Amz links. So now on order.
This doesn't protect the cable modem, but the recommendation was to switch to an inline gas surge protector like the HAM guys use on their setups:
https://www.amazon.com/dp/B07GBW7XKX
I am sorry for your loss. I know your Hubitat was a dear friend and I hope you can move on from this tragedy.
As for the coax lightning arrester, what are you tying that into?
Standard electrical code says it should go into an eight-foot copper rod (half inch??) driven straight into the ground connected with 10 gauge preferably. Someone was recently telling me that around here the ground is often quite dry and standard is not enough. Engineers and electricians who want meet grounding requirements can end up using several rods and deeper/longer ones. My solution was to put it near the hose bib so that the drip drip drip from leaky hoses keeps adding some moisture to the ground.
When I built the house in the 80s I had two 5/8" x 8 ft copper-clad ground rods driven - one near the service entrance meter and one near the communications cabinet. (The water table here is about 24" from the surface much of the time and I've never seen it lower than 36".) I can get to them with #10 and a little effort.
BTW, the electrical engineers & commercial installers on Hubitat say that the Code now calls for multiple grounds rods to be bonded together --probably with #6 copper wire.
I have almost everything working again. What a monumental PITA.
>everything working again.
When I first went with Hubitat I opted for zwave devices (longer range, more types of devices available). The forums kept buzzing about maintenance issues with zwave while the zigbee guys had little of that. I tossed out all but 3 or 4 zwave devices that had no zigbee replacement and went zigbee. THAT turned out to be a great move and the main reason my system just didn't fail me.
Still, having to replace the 3yo hubitat console has some benefits. I cleaned up the installation in general, got rid of the 4 legacy zwave devices, and trimmed a few zigbee devices that I rarely used. It also forced me to reacquaint myself with the network and the array of installed devices --this is a problem when a system is solid and runs well autonomously. I also documented and built a cheat sheet re how to pair each type of device to the hub. Each brand seems to have its own secret handshake and re-pairing is often slightly different than initial setup.
That said, while setting up the replacement hub wasn't too hard, researching the secret code of the devices --many now obsolete and the manufacturers out of business-- and then re-pairing them was a slow process. And I have about 80 of them (a small-to-medium sized smarthome network kept that way on purpose).
The good news is that Hubitat had processes for backing up the database and the programs I had built with it. Better yet, I had been keeping fairly current backup copies on my laptop. Otherwise, I'd have lost weeks of work.
Quote from: rcjordan on May 11, 2022, 10:58:10 PM
Code now calls for multiple grounds rods to be bonded together --probably with #6 copper wire.
It depends on your soil. The code says you must have an electrode and a supplemental electrode unless you get a resistance to the ground of 25 ohms or less with a single electrode.
This is apparently for MN, but I think it's just the standard NEC code.
https://up.codes/viewer/minnesota/nfpa-70-2020/chapter/2/wiring-and-protection#250.53_(A)_(2)
Nice write-up
https://www.ecmweb.com/content/article/20886735/bringing-grounding-down-to-earth
That article notes that since two rods meets code regardless of resistance, many contractors just drive two rods and call it done and don't bother to measure resistance. I guess what I heard is that some contractors *will* measure resistance and they will try to hit the 25-ohm threshold no matter how many rods it takes. A builder friend said an electrician he worked with said that in very dry soil they will bury huge arrays of rods to try to meet that standard (double digit numbers).
For a 200 amp service, the connecting wire needs to be gauge 6 it turns out. In my case, I'm just trying to ground coax cable with can only carry so much current. I'm reasonably sure (???) that by the time I max out a 10-gauge wire, the coax cable will have melted and caught fire. I think the point of the ground is to give the current somewhere to go other than the sensitive equipment.
I *think* that's different from grounding a 200-amp service, but all this has me thinking about driving another rod and running heavier wire.
In my case, this is to ground a cell booster antenna 100-feet up a tree. The current will travel through 100-feet of 0.5 ohm coax before it even hits the ground connector. Aside from burning down the tree or house, right after the ground connecter it goes into a $700 booster that I'd prefer not to fry.
The multi-point grounding concern is that the surge will jump to the building's neutral wires as it's the least resistance. Then all hell breaks loose.
That would be very hard in my scenario - it would have to run 100 feet of coax, fail to ground at the grounding junction, find it's way through a sensitive booster, then go into the neutral wire. That does not seem like a path of least resistance
>recommendation was to switch to an inline gas surge protector like the HAM guys use
I wanted to get the network up and fairly stable before I started installing the lightning protection. I've just installed the gas tube surge protector where the ISP's incoming coax comes into the comm cabinet. So far, so good. No noticeable speed degradation, but I'm not the streamer in the family. We'll see what happens when she cranks up Netflix along with her Ipad plus her phone --all at once, of course.
>when she cranks up Netflix
No problems.
I just installed the optical surge protector setup. No problems so far with that, either.
To wrap this up, I installed an ethernet surge protector as a secondary. It's installed between the network switch and the long cat5 leg going to the hubitat. Again, no noticeable degradation of service.
Ethernet Surge Protector
https://www.amazon.com/gp/product/B07GBLFFNK
Can we back up just a bit?
First question: what is the danger point you're protecting from? In other words, in my case it's an antenna on a coax cable high in a tree.
Second question: starting from that point, what is the path from there to you sensitive electronics? In other words, it comes down the cable or through the exterior power feed or whatever, and what surge suppressors does it hit on each line and what order?
Not asking for excruciating detail. I'm just trying to visualize the system (and thinking about where/how I might want to upgrade mine).
>danger point
Incoming coax from the cable company. It is buried in my neighborhood & front yard, but there are exposed terminal blocks. The surge seems to follow the coax through the comm cabinet entry point in the garage. The cable modem gets fried. I've lost several over the years without damage to the router which was the next device upstream. Cable modems are the least expensive equipment --$100 at Walmart, so sorta sacrificial.
>and what surge suppressors does it hit on each line and what order?
==> 75ohm inline gas tube coax surge protector HERE
1 - cable modem
==> optical surge protector setup HERE
2 - router
3- network switch
==> unpowered ethernet surge protector HERE
4 - hubitat
Thanks. I realized that after my big UPS failed yet again and I no longer need it anyway because of the whole-house battery, the surge suppression function of the UPS has also been removed from my setup.
Lots of stuff in your list I've never heard of - gas chamber surge suppressor, optical surge suppressor. I'm going to need to do some research and up my game.
>UPS
There are several posts saying that having a UPS doesn't stop it. Several of us have had it come in over the coax.
Copy/paste from others:
Naive question, but would putting your networking gear and HE behind a UPS potentially prevented this?
For me, the strike traveled along the Cat 5e and fried the ports on the switches and motherboard. UPS wasn't involved as it came in the other end. All the gear that died was on different UPSes, none-the-less.
Yes, the coax is a problem. The UPS has surge-protected ethernet ports, so the Cat-5 was running through surge protection.
The order was
- satellite dish >> coax >> modem >> Cat5 >> surge protector >> all the other stuff
But I'm now thinking about the cell booster, and the order there is
- antenna 100 feet up a tree >> 100 feet of 0.5 ohm coax >> lightning ground >> sensitive equipment
That has me concerned. The saving grace is that we are surrounded by ridges and 140+ foot trees. And as I write that, I think I should share some of this info with friends who are building on a ridge top
Is lightning a frequent problem there? In the spring, we probably average 2 thunderstorms a week. In the summer, up that to 3. I once read that we're in the 2nd-most prolific lightning belt. The storms are not usually uber-violent like those in the prairie states.
>trees
My neighbor's house was originally built as a 'tree house' --up on stilts and nestled in a grove with numberous large trees just inches off the deck and deck pole lights. That house was my lightning protection for a couple of decades. About twice a year, lightning would strike one of those nearby trees, jump to the deck lighting, then usually fry something inside. Subsequent owners have cut the trees that are closeby --and now my yard gets hit. Indications are it is most often the Spectrum cable network that gets hit, but it might jump there after running down a tree.
Lightning is rare here. First off, summer thunderstorms are mostly at higher altitude (above about 8000 feet and we're down at 5700). Down here, we might hear thunder a couple times a year.
Second, we are surrounded by trees and ridges and we're fairly low. But even up on the ridges near me (6000 feet) lightning strikes are not especially common.
If they were more somewhat common here, we would have burned down a long time ago. If they were a lot more common here, we would have a lot lower fuel load and probably be mostly safe.
>2nd highest
Having one now.
https://www.blitzortung.org/en/live_lightning_maps.php?map=33
That's a pretty wild map.
We get the occasional dry lightning storm that passes through, sets up to 1000 fires across the states (I think 1200 fires in one storms some years back) and then moves on.
Lightning strikes resident inside home, PA officials say
https://www.charlotteobserver.com/news/nation-world/national/article261514887.html