CAT5 surge/lightning strike protection recommendations?
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market. Anyone have recommendations (tested/practical is best :-)? The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons? Thanks, Adi
Fiber would be my choice. Not only will it solve the lightening strike problem; you will not have to worry about ground potentials being different on each side of the cable run. James Routing and Security Administrator At the Santa Fe Office: Cyber Mesa Telecom jamesh@cybermesa.com noc@cybermesa.com http://www.cybermesa.com/ContactCM (505) 795-7101
On Tue, 13 Sep 2005, R.P. Aditya wrote:
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports.
Don't do that, then.
I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations (tested/practical is best :-)?
Use the cat5 as a pull rope, install fiber.
The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons?
Seriously, this is a battle against Mother Nature that you aren't going to win. Differences in ground potential as well as induced currents into the UTP will continue to cause equipment failure as well as possibly kill you or someone else. -- Jay Hennigan - CCIE #7880 - Network Administration - jay@west.net WestNet: Connecting you to the planet. 805 884-6323 WB6RDV NetLojix Communications, Inc. - http://www.netlojix.com/
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations (tested/practical is best :-)?
The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons?
Adi, Is there a reason that your between-building runs aren't being done with Fibre? It being non-conductive is one immediate advantage.... Also if your grounding is inadequate you may like to take a squiz at the ISO or TIA Standards as they pertain to cabling. In NZ we have a variety of standards which all point back to ISO, the ANSI equivalents are TIA/EIA 568-B (Cabling), TIA/EIA-569-A (Pathways and Spaces) and TIA/EIA-607-A (Electrical Wiring, relevant as it pertains to Earthing etc). Even for short runs, If I need to run between buildings externally I won't even look at copper. Mark.
On Wed, 14 Sep 2005 12:24:39 +1200 (NZST) "Mark Foster" <blakjak@blakjak.net> wrote:
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations (tested/practical is best :-)?
The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons?
Adi,
Is there a reason that your between-building runs aren't being done with Fibre? It being non-conductive is one immediate advantage....
I would agree with Mark. Even buried copper can make an excellent guide for lightning to come right into your equipment, and it can only be isolated so much. (Remember, the electrical potential of the ground can vary over a cable run, and will vary if there are elevation changes.) Fiber is the way to go. Regards Marshall Eubanks
Also if your grounding is inadequate you may like to take a squiz at the ISO or TIA Standards as they pertain to cabling. In NZ we have a variety of standards which all point back to ISO, the ANSI equivalents are TIA/EIA 568-B (Cabling), TIA/EIA-569-A (Pathways and Spaces) and TIA/EIA-607-A (Electrical Wiring, relevant as it pertains to Earthing etc).
Even for short runs, If I need to run between buildings externally I won't even look at copper.
Mark.
R.P. Aditya wrote:
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations (tested/practical is best :-)?
The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons?
Thanks, Adi
I'll go with the fiber recommendation but that's not what you asked :-) We use quite a few of the Motorola 300SS surge suppressors. They are made for use with Motorola's fixed wireless Internet platform and go on the Ethernet cable before it enters the building. They do a good job of protecting the ports on near misses. Direct strikes and they are toast along with anything attached to them - but that's just the way it goes :-) http://www.tessco.com/products/headerProductSearch.do?searchType=1&searchText=300ss&searchField=1 -- Mark Radabaugh Amplex mark@amplex.net 419.837.5015
Speaking on Deep Background, the Press Secretary whispered:
I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations
A) Don't. B) Don't C)... Surge protectors can not protect you from ground differential issues. Your answer is 1) Pull fiber with that CAT5 pullrope. 2) If you REALLY, REALLY can't.... Put a fiber transceiver in building A. At least 10 foot away, put in a 2nd transceiver and connect THAT to the CAT5 going to building B. Connect A & B wallwarts to different breakers, with surge protectors....and stock spares.o {Extra help; power B from a small hospital-grade isolation transformer -- low leakage, hi breakdown voltage.} Hopefully, you'll merely lose transceivers & wallwarts on the B-side, with nothing in building B being zorched. -- A host is a host from coast to coast.................wb8foz@nrk.com & no one will talk to a host that's close........[v].(301) 56-LINUX Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433
David Lesher wrote:
Surge protectors can not protect you from ground differential issues.
True enough - but 10/100 Ethernet is normally isolated by the transformer on the Ethernet transceiver. AFAIK there is not a connection between the signal lines and ground. Isolation is 1500V for the magnetics I checked. Off course all bets are off when lightning strikes since the voltage tends to be just a tad higher than 1500 volts..... Mark Radabaugh
On Tue, 13 Sep 2005, David Lesher wrote:
Put a fiber transceiver in building A. At least 10 foot away, put in a 2nd transceiver and connect THAT to the CAT5 going to building B. Connect A & B wallwarts to different breakers, with surge protectors....and stock spares.o
That's an amazingly expensive optoisolator. Seriously, though, that's exactly what you're describing, and about what I'd suggest in a no-other-option scenario -- but if it's possible to pull fiber through the conduits, it would probably be far less expensive long term, or even medium term if the physical fiber spools can be bought cheaply enough. -- -- Todd Vierling <tv@duh.org> <tv@pobox.com> <todd@vierling.name>
As everyone else has said, fiber is best, but if that is not an option... We have had good luck using these: http://www.hyperlinktech.com/web/hgln_cat6.php Trancievers will work as well, but that is a more expensive option. Nothing is going to protect you from a direct strike. Jerry
Todd Vierling <tv@duh.org> writes:
Seriously, though, that's exactly what you're describing, and about what I'd suggest in a no-other-option scenario -- but if it's possible to pull fiber through the conduits, it would probably be far less expensive long term, or even medium term if the physical fiber spools can be bought cheaply enough.
For those who haven't priced the stuff lately, in spools of 1000' the per-foot prices of 2-strand MM tight buffered fiber suitable for pulling in conduits like he (hopefully) has tends to be price-competitive with cat5 on a per-foot basis. Extra strands are cheap; the pricey part of fiber is the jacket and strength members; even super-pure glass is not that expensive overall. The expensive parts in the equation turn out to be the termination trays and connectors. ---Rob
In message <87ll1zn81b.fsf@valhalla.seastrom.com>, "Robert E.Seastrom" writes:
Todd Vierling <tv@duh.org> writes:
Seriously, though, that's exactly what you're describing, and about what I'd suggest in a no-other-option scenario -- but if it's possible to pull fiber through the conduits, it would probably be far less expensive long term, or even medium term if the physical fiber spools can be bought cheaply enough.
For those who haven't priced the stuff lately, in spools of 1000' the per-foot prices of 2-strand MM tight buffered fiber suitable for pulling in conduits like he (hopefully) has tends to be price-competitive with cat5 on a per-foot basis. Extra strands are cheap; the pricey part of fiber is the jacket and strength members; even super-pure glass is not that expensive overall.
The expensive parts in the equation turn out to be the termination trays and connectors.
Also the labor of pulling it, when there's already something in the (shudder) ground. My direct experience with running long-distance underground cable is dated -- let's put it like this; we were dealing with RS-232 -- but the countermeasures to a direct strike on copper cables don't seem to have improved nearly enough... --Steven M. Bellovin, http://www.cs.columbia.edu/~smb
On Wed, 14 Sep 2005 08:44:16 -0400 "Steven M. Bellovin" <smb@cs.columbia.edu> wrote:
In message <87ll1zn81b.fsf@valhalla.seastrom.com>, "Robert E.Seastrom" writes:
Todd Vierling <tv@duh.org> writes:
Seriously, though, that's exactly what you're describing, and about what I'd suggest in a no-other-option scenario -- but if it's possible to pull fiber through the conduits, it would probably be far less expensive long term, or even medium term if the physical fiber spools can be bought cheaply enough.
For those who haven't priced the stuff lately, in spools of 1000' the per-foot prices of 2-strand MM tight buffered fiber suitable for pulling in conduits like he (hopefully) has tends to be price-competitive with cat5 on a per-foot basis. Extra strands are cheap; the pricey part of fiber is the jacket and strength members; even super-pure glass is not that expensive overall.
The expensive parts in the equation turn out to be the termination trays and connectors.
Also the labor of pulling it, when there's already something in the (shudder) ground.
My direct experience with running long-distance underground cable is dated -- let's put it like this; we were dealing with RS-232 -- but the countermeasures to a direct strike on copper cables don't seem to have improved nearly enough...
I don't think they will... tens of megavolts is hard to protect against. This depends a little on where you are. I have experience with cable runs in Southern Florida (where lightning strikes can occur daily), West Virginia and Virginia (with strikes common) and Hawaii (where they don't seem to be as frequent). The cable may be in the ground, but it is connected to stuff at either end which isn't, and given the potential differences that occur in the natural environment (~ 150 volts per meter of altitude), this means that cable runs can act like lightning rods connected directly to your network gear. So my jaded perspective is that you WILL get hit if you connect buildings with copper, and you WILL NOT like it. Since this can be entirely mitigated through the use of fiber, use fiber if you possibly can. Regards
--Steven M. Bellovin, http://www.cs.columbia.edu/~smb
In message <web-2973284@multicasttech.com>, "Marshall Eubanks" writes:
My direct experience with running long-distance underground cable is dated -- let's put it like this; we were dealing with RS-232 -- but the countermeasures to a direct strike on copper cables don't seem to have improved nearly enough...
I don't think they will... tens of megavolts is hard to protect against.
This depends a little on where you are. I have experience with cable runs in Southern Florida (where lightning strikes can occur daily), West Virginia and Virginia (with strikes common) and Hawaii (where they don't seem to be as frequent). The cable may be in the ground, but it is connected to stuff at either end which isn't, and given the potential di fferences that occur in the natural environment (~ 150 volts per meter of altitude), this mea ns that cable runs can act like lightning rods connected directly to your network gear.
So my jaded perspective is that you WILL get hit if you connect buildings with copper, and you WILL NOT like it. Since this can be entirely mitigated th rough the use of fiber, use fiber if you possibly can.
Right. When I lived in North Carolina, there was a ground strike close to my apartment. It tripped some circuit breakers, burned out some light bulbs, and fried the cable TV box, the balun, and the RF input on my TV. And what regularly happened to our computers and terminal gear in the CS department was scary.
Anyone have recommendations (tested/practical is best :-)? First bond the cable shield to earth at the entry point in both buildings. Second use a Tripplite DNET-1 at both buildings. make sure you have a single
On Tuesday 13 September 2005 03:24 pm, R.P. Aditya wrote: point ground terminal that everything it bonded to and then that is grounded. I am in the chicago area, and have towers and locations that have ethernet cables running up to radios. We use this meathod, and don't have problems. Worst we had after last night's T-storm was a site with out power :-) -- Bryan Fields Chief RF Engineer/Partner illiana.net 219-306-1805
On Tue, 13 Sep 2005 20:24:51 +0000, "R.P. Aditya" <aditya@grot.org> said: I have a bunch of cat5 buried about 1 ft below the surface connecting multiple buildings on a campus (short runs) and lightning strikes nearby have caused surges along one or more of the cables and burnt out switch ports. I would like to protect the switch ports -- there seem to be lots of products on the market.
Anyone have recommendations (tested/practical is best :-)?
The APC Protectnet PNET1 and PRM24 seem quite nice and not too expensive -- if they work....pros? cons?
Thanks to everyone who replied on- and off-list. The installation in question is in a condo development and was done by licensed electricians and the residents were lead to believe that it was code compliant. The cat5 cabling is double-sheathed with a moisture barrier. As you can well imagine, the residents are very cost-concious. My preference is that fiber be run in conduits, however even running cat5 in grounded conduit is a big hassle as it will involve cutting across pavement etc. (I fully appreciate the danger from potential difference between buildings and copper being a good conducter etc., but I had to ask a leading question in order to document the problem such that sufficient notice would be paid by the residents -- I believe I have that now). The short-term solution seems to be using the APC PNET1s/Tripplite DNET1/etc. in each unit and tying them to the water main as an inexpensive, immediate step while funds are allocated for conduit, labor etc.. Thanks, Adi
In message <20050914184430.GA54042@mighty.grot.org>, Aditya writes:
The short-term solution seems to be using the APC PNET1s/Tripplite DNET1/etc. in each unit and tying them to the water main as an inexpensive, immediate step while funds are allocated for conduit, labor etc..
If I recall correctly, the National Electrical Code was change about a dozen years ago to bar grounding to water pipe unless it's within about 6' of where the pipe enters the house. --Steven M. Bellovin, http://www.cs.columbia.edu/~smb
Steven M. Bellovin wrote:
In message <20050914184430.GA54042@mighty.grot.org>, Aditya writes:
The short-term solution seems to be using the APC PNET1s/Tripplite DNET1/etc. in each unit and tying them to the water main as an inexpensive, immediate step while funds are allocated for conduit, labor etc..
If I recall correctly, the National Electrical Code was change about a dozen years ago to bar grounding to water pipe unless it's within about 6' of where the pipe enters the house.
--Steven M. Bellovin, http://www.cs.columbia.edu/~smb
The change was in the 1993 NEC and restricts using water pipe as a ground to being within 5 feet of the point of entrance to the building. There is an additional requirement that the water pipe must be metallic and have a minimum of 10' buried in the ground. This can be hard to verify on existing installations due to the common use of plastic water pipe by utilities in recent years. The 5' from the point of entrance is also due to the common use of plastic water pipe in buildings. -- Mark Radabaugh Amplex mark@amplex.net 419.837.5015
participants (13)
-
Aditya -
Bryan Fields -
David Lesher -
james edwards -
Jay Hennigan -
Jerry Eyers -
Mark Foster -
Mark Radabaugh -
Marshall Eubanks -
R.P. Aditya
-
Robert E.Seastrom -
Steven M. Bellovin -
Todd Vierling