| POE and lightning Well we took a hit last night, lost 5 tranzeo tr5a Plus radios and one ceragon backhaul. All were POE, all using shielded cat5 and all had poe surge suppressors on each end. Surge suppressors on the top were all blown, so were the ethernet ports on the radios directly behind them. We took a similiar hit on this same tower last year, blowing the ethernet ports on nearly every POE radio on that tower(all but one). I then doubled up on the surge suppressors hoping to stop that. IT didn't.
Since this seems not to be working would this help.
Metal Conduit with the POE cables within the conduit. I can get some flexible aluminum conduit quite cheap at lowes/home depot. And am considering re-running all new cat5 cables within the flexible aluminum conduit. My thinking is that the cat5 cables are not getting hit, just something close to where they are all run, and the proximity is allowing some of the charge to go through the cat5 cables. Do you all think this would help??? If so I'll be on it sometime next week.
BM |
|
robbinPremium,MVM
join:2000-09-21
Leander, TX
kudos:1 | How well is the tower grounded? |
|
|
|
2 edits | Quite Well.
All guys are grounded, three 15 footers at the base bonded to the (2) electrical grounds(The headend building has two drops from two different electric circuits).
I have other gear on the tower, only the POE equip got hurt.
Ohh, yeah it's a 300 foot rohn 45 with POE gear mounted at 140,160,225,275 and 290feet. The gear mounted at 225 feet survived.
BM |
|
robbinPremium,MVM
join:2000-09-21
Leander, TX
kudos:1 | Have you tested the ground system? What is the ohm? |
|
| I don't have an ohm meter to check that. However I don't believe it to be a ground issue. An 8 footground around here in typical dirt will always pull 5 ohms(at least thats what the electricians and they guy who works for the telco told me(they have an engineer for tower sites who makes sure grounding is up to par, before they will hook a T1/DS3 up to a tower site).
All the cat5 cable runs beside (3) 1 & 5/8" Heliax Cables, the radios attached to those 300 foot sections have yet to be hurt. The Heliax is grounded at the top and bottom, and there are polyphaser lightning arrestors right before the radio.
There is also some gear up top in an enclosure case that runs lmr to the antenna, they have yet to be hit in over 4 years. The switch up there however gets hit atleast once a year.
BM |
|
robbinPremium,MVM
join:2000-09-21
Leander, TX
kudos:1 | Sounds like you have covered the bases pretty well. I have heard that Tranzeo is vulnerable to lightning. Just to be clear -- all of the lightning protectors are polyphaser? How are they grounded to the tower (welded stud, clamp, ground wire to base)? |
|
1 edit | The poe suppressors are all Altelicon. I don't recall the model #, but you have to manually wire them in. It's a bit of a pain when not on the ground and quite time consuming no matter where you are.
All POE suppressors are clamped to a bare(and well polished) part of the tower using 10 ga wire. The ones at the base are attached with 10ga wire to a bus bar with a 4ga wire that is attached to the entire grounding system. (2) of the surge suppressors at the base were blown.
A ground up the tower from my understanding would be quite useless unless it were a true earth ground strap. I would assume the resistance of a rohn 45 tower is quite low. There is a heck of a lot of steel in that thing. It's probably my favorite tower to climb, though I'd have rather not climbed it twice today. Will be climbing it and a water tower tommorrow(wind blew a P2P link out of alignment at the water tower, must not have had it as tight as I thought I did, though I could not personally move it).
BM |
|
| I'd be curious as to how your STP runs are terminated. Are you using STP connectors on both ends of the span? Are you connecting the bottom drain to your master buss bar in the shack? |
|
1 edit | Yes, the drain is attached at the bottom.
Does anyone think the conduit might help?
BM |
|
| I do not know the answer but I bet John Galt does.... maybe he will wander in here.
I would say its going to be a royal PITA to do conduit. You are going to need to use hangers to hang the conduit up the tower, then you are going to need to bond it at regular intervals.
I'd say it is possible that the pulse from the strike that is generated could be inducting into the cat5, but with it being STP, heck you have me.....
--
"No job is so important, and no service is so urgent that we cannot take the time to do it safely."
-- AT&T
--Safety One Tower Rescue Certified |
|
| The conduit won't be that big a deal. It will take me about 2 days. I'd rather spend two days on it now then go through this crap. Oh yeah the radios we had shipped overnight are stuck in customs!!! 3 hours away, been there since 2:30 in the morning. If they will let us go down there and pick it up I won't get it hung and finished until 8 tonight.
My spare gear decided it no longer wanted to work. The replacement backhaul had a bad radio on one side......Freakin Lightning.
I now have a temp 2.4 backhaul to the tower. Thank god I had some delib 2.4 gear in.
It's either conduit or Place Two back to back surge suppressors at the top. All of the radios still powered up, just would not pass trafic through the ehternet ports.
BM |
|
 John GaltForward, MarchPremium
join:2004-09-30
Happy Camp
kudos:3
1 edit | Pics of the grounding?
Actually, pics of any part of the installation...more is better.
--
A is A |
|
lutfulPremium
join:2005-06-16
Ottawa, ON Reviews:
 ·TekSavvy DSL
1 edit | reply to milbrath
Altelicon/Hyperlink |
1. Most PoE surge protectors clamp PoE PWR and GND wires to surge ground at 56V max using MOVs. Tranzeo and other "less than 48V" PoE may die at 56V.
2. Many PoE surge protectors clamp Ethernet pos and neg pairs to surge ground at 15V max using silcon avalanche diodes or equivalent fast reaction devices.
Sadly most modern Ethernet chipsets accept 3V maximum and will die at 15V within microseconds.
3. Most RF surge protectors clamp center conductor to 90V using slow gas discharge tubes (GDT) . The voltage could rise upto 300V before GDT triggers and kill many radios.
Unless every single surge entry-point ( PoE PWR and GND pairs, Ethernet pos/neg pairs and RF center conductor and shied ) is properly protected, radio PCB will encounter MASSIVE ground potential differences during thunderstorm.
Bond all surge protector grounds together and also to significant metal of radio enclosure. Bond to tower body with short wide strap.
Edit: Found Altelicon/Hyperlink PoE protector schematic. It clamps PWR/GND at 58V and data at 15V. Both are higher than Tranzeo and many other PCB's max input specification. Citel (resold by Demarctech and Deliberant) are a bit better in spec, but still higher than manyh indoor radio specs. |
|
public
join:2002-01-19
Santa Clara, CA | reply to milbrath
said by milbrath:Does anyone think the conduit might help? Is is like adding another shield to the cable.
But if your protectors burn up, you also need higher surge energy rated protectors. Sealed spark gaps have high surge energy rating and low capacitance needed on data lines. |
|
lutfulPremium
join:2005-06-16
Ottawa, ON | Do you know any for PoE (60V max) or Ethernet (6V max) ratings? Epcos spark gaps are all rated for AC line voltages. I recall DEHN uses them. |
|
public
join:2002-01-19
Santa Clara, CA | said by lutful:Do you know any for PoE (60V max) or Ethernet (6V max) ratings? Epcos spark gaps are all rated for AC line voltages. I recall DEHN uses them. Spark gaps are usually constructed for 50V or higher trigger voltage. But ethernet transformers have 100V isolation so even easy to get 90V gas filled spark gap will provide some protection. Okaya, Siemens, Littlefuse, etc...
Power pairs do not need low capacitance, any protector with enough surge energy rating will work.
There are also multiple type protector device assemblies available off the shelf, for example
»www.lightningprotectioncor.com/w···tors.htm
Obviously all protectors at the AP must be referenced to the same ground. |
|
lutfulPremium
join:2005-06-16
Ottawa, ON Reviews:
·TekSavvy DSL
1 edit | said by public :
easy to get 90V gas filled spark gap will provide some protection Are you talking about traditional gas discharge tubes?
GDTs are not really suitable for Ethernet protection all by themselves - you will need to parallel ultra-fast silicon devices like these: »www.semtech.com/mediacenter/prin···t_ad.pdf
I was thinking you meant those new "triggered" spark gaps with lower "let through" voltage. Here is one example device:
»www.erico.com/public/library/fep···_P26.pdf |
|
 harvSkiPremium
join:2004-03-09
London, UK | reply to milbrath
Clamping an aluminium conduit to a steel tower may cause electrolysis problems - i.e. the tower may rust faster. If the tower is galvanised I think the aluminium will strip the zinc off the steel faster than if there was no aluminium about.
Bad news about the PoE surge arrestors - I'm planning to install some on my kit soon. |
|
| reply to milbrath
Of course, Me, not being an engineer would start to wonder why you have been hit twice. This in some ways goes to show you can only do so much to protect from a direct hit. But mainly in my mind, I would really start to question the towers grounding. If that tower got hit often, is it possible that the grounding of that tower is weak. Allowing to much resistance for the lightning hit to dissipate quickly.
I'm not griping at you Milbrath, and I feel for ya. That totally sucks loosing equipment twice. And a fair amount of equipment at that. You have certainly put forth the effort to protect your equipment from this. More so than I for sure. I guess some just have bad luck, and others even worse luck  |
|
| I've doubled up the surge suppressors at the top for now(1 at the bottom 2 right before the AP's). I am going to hang new cables with each cable enclosed in conduit next week. All my gear was still powering up and clients were still associating with the radios, just no ethernet activity on the tower side. Just now got every new radio re-hung. So none of my gear actually got hit. It was cold and windy at 6 am this morning. I feel for the yankees in here working on towers.
I may ask one of the electric company's employees to test the ohm rating of the grouning system.
BM |
|
