>>opto-isolation is that when it comes to lightning, the voltages can be
enormous--a few thousand volts of isolation is "nothing.
Usually, you run it in metal conduit, so lightning is not an issue. It's
designed to cope with differences in building ground potential.
>>Better is to locally ground both ends of the Ethernet via Ethernet surge
protectors
Well... that doesn't solve the ground potential difference, which is usually
a bigger issue than lightning. Not a bad thing to add, however.
>>fiber all-the-way gives enough protection.
Yes, that is by far the best.
>>Compared to dry sand, the water apparently was a better conductor.
Pure water is not a good conductor, but ground water is extremely
conductive. Dry sand is mostly silica - a great insulator. Cycloaliphatic
epoxy insulators are typically 60-70% silica.
>>Fla Power (or TECO?), a number of decades ago tested underground power
lines (east of Bradenton, as I
remember) and abandoned the idea as current from lightning was putting holes
in the jacket (pvc & metal shield).
Early concentric neutral underground power cable has serious problems with
insulation degradation due to an effect known as "treeing". Flaws would
develop in the insulation due to the extremely high electrical stress,
lowering the basic insulation level (BIL) until a spike (lightning, or even
the operation of a switch or capacitor bank) would punch right through. It
took so long to restore service, that many utilities gave up.
>>What I have been doing is to bury a bare copper ground wire...
Hmmm.... Since it in the ground, which is where lightning wants to go, I'm
not sure its really doing anything of value. Can't hurt. It will corrode
over time, as electric utilities. On second thought, it might be a good
thing in sandy soil, during the dry season, since dry sand is a poor
conductor.
Building to building, I like optical isolation in grounded metal conduit. I
really hate wireless, because of all the security issues. But we all have
our opinions...
Ken Elliott
=====================
-----Original Message-----
From: slug@nks.net [mailto:slug@nks.net] On Behalf Of Donald E Haselwood
Sent: Tuesday, July 20, 2004 1:49 PM
To: slug@nks.net
Subject: {SPAM?} RE: [SLUG] Using old ethernet wiring?
The problem with opto-isolation is that when it comes to lightning, the
voltages can be enormous--a few thousand volts of isolation is "nothing."
* Better is to locally ground both ends of the ethernet via ethernet surge
protectors (about $25), though fiber all-the-way gives enough protection.
Regarding PVC: My mother-in-law had lightning strike a big pine tree. One
path it took was into the water line, through the PVC. Compared to dry
sand, the water apparently was a better conductor. I would expect lightning
to "find" wire inside buried PVC. Fla Power (or TECO?), a number of decades
ago tested underground power lines (east of Bradenton, as I
remember) and abandoned the idea as current from lightning was putting holes
in the jacket (pvc & metal shield). What I have been doing is to bury a
bare copper ground wire #6 or #8 (cheaper and should be quite
adequate) along side the PVC pipe carrying the wires; the outside wire is
tied to the ground system of course. Any ground current will follow this
wire, and it is much less likely that enough voltage difference will exist
to punch through the PVC.
Using wireless really gets around a difficult problem for us folk in central
FL.
Donald Haselwood
* The average lightning stroke has a peak current of about 18,000 amps; 90%
of the strokes will be less than 100,000amps. A single ground rod has a
resistance of about 30 ohms, so on *average* you get a peak voltage of about
1/2 million volts, and 10% of the time over 3 million volts--that's with the
system grounded! Obviously, the discharge finds other paths.
Even more important than resistance is the inductance. The rate-of-rise of
current is upwards of 1E10 amps/sec, so the voltage across just 10 feet of
#6 ground wire is around 10,000 volts. This means that if two pieces of
equipment are connected to the ground wire, but 10 feet apart, an ethernet
link between them is got to deal with voltages in 1,000 to 10,000 volt range
(i.e. blasted NIC, or worse).
At 09:12 PM 7/19/04, you wrote:
> >>Lightning will take any path, conductive or not, particularly if
> >>there's
>something solid (like PVC) to guide it.
>
>Well, not exactly... Lightning is electricity. It requires a
>conductive path. It passes through the air when the electrical stress
>caused the nitrogen to break down and ionize (associated with partial
>discharge, or "corona"). I've never seen it hit PVC, unless it was
>badly contaminated or contained conductors. I could be wrong, but I
>used to do a lot of high voltage testing, and designed 150 kV insulators
and bushings.
>
>HP used to make optical isolators for hooking Ethernet between
>buildings, that let you run twisted pair. You could also tie all the
>buildings to a common ground point, but that's way too much work.
>
>Ken Elliott
>
>=====================
>-----Original Message-----
>From: slug@nks.net [mailto:slug@nks.net] On Behalf Of Paul M Foster
>Sent: Monday, July 19, 2004 1:22 AM
>To: slug@nks.net
>Subject: Re: [SLUG] Using old ethernet wiring?
>
>On Sun, Jul 18, 2004 at 11:27:30PM -0400, Eben King wrote:
>
> > On Sat, 17 Jul 2004 wchast@utilpart.com wrote:
> >
>
><snip>
>
> > > I used this same setup to pull 300 feet of cord through PVC piping
> > > between two houses then pulled cat5 through it. Needed the lub
> > > though as it got a bit hard about 3/4 of the way through. But at
> > > least I have the two buildings linked.
> >
> > Shouldn't put something conductive between buildings. They might
> > have different ground potential (it works, thus disproving that), or
> > lightning could strike nearby, frying one or both ends and attached
> > equipment. You should use fiber or radio.
>
>That's why you run a ground wire along with the cable and tie it to the
>grounding system at both ends. ;-} (I know, there could still be
>differences, even if the ground is a continuous loop between
>buildings.) Or you could bury it (best idea, but still not perfect).
>
>And if lightning strikes nearby, it won't matter if it's conductive or not.
>Lightning will take any path, conductive or not, particularly if
>there's something solid (like PVC) to guide it. Conductive solids only
>present a more enticing target for lightning.
>
>Paul
>-----------------------------------------------------------------------
>This list is provided as an unmoderated internet service by Networked
>Knowledge Systems (NKS). Views and opinions expressed in messages
>posted are those of the author and do not necessarily reflect the
>official policy or position of NKS or any of its employees.
>
>-----------------------------------------------------------------------
>This list is provided as an unmoderated internet service by Networked
>Knowledge Systems (NKS). Views and opinions expressed in messages
>posted are those of the author and do not necessarily reflect the
>official policy or position of NKS or any of its employees.
-----------------------------------------------------------------------
This list is provided as an unmoderated internet service by Networked
Knowledge Systems (NKS). Views and opinions expressed in messages posted
are those of the author and do not necessarily reflect the official policy
or position of NKS or any of its employees.
-----------------------------------------------------------------------
This list is provided as an unmoderated internet service by Networked
Knowledge Systems (NKS). Views and opinions expressed in messages
posted are those of the author and do not necessarily reflect the
official policy or position of NKS or any of its employees.
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 20:40:54 EDT