Jim
Your are correct in all of your comments and I agree fully. I did not go to
that level of detail as I was trying to keep my comments relatively simple.
The reason I mentioned that non UL listed devices may be acceptable for
transmission lines is the exact item you mentioned. Few are UL listed,
therefore difficult to find a suitable device at a reasonable price.
Yes, UL only lets us know that it is not likely to catch fire and burn
something up. Has little to do with functionality as you point out.
Series type AC SPDs are preferred but the size is generally prohibitive.
They are rather large for a 3 phase 400 amp service. Parallel devices
properly installed with short, generally straight conductors that have been
installed at the service entrance or on a separately derived system are
relatively effective. Series devices for telephone type circuits are most
desirable and are reasonable. You are correct about the MOV failures. They
should be examined or better, replaced, after any significant transient
event. They will degrade over time just from the spikes created by shutting
of motors or any inductive load.
The average wave form is thought to be 8/20 but various events have
indicated it is all across the board but with a rather fast rise time. In
any event it is generally broad band until "tuned" by some circuit
components. I have seen the energy go right through VHF and UHF cavity
filters.
I typically work with the public safety folks who demand a high level of
protection to ensure reliable service. A few thousand dollars is nothing
when you have a 1 Million dollar investment. But if you desire the best it
can get expensive.
Thanks for your comments.
Regards
Bruce
-----Original Message-----
From: jimlux [mailto:jimlux@earthlink.net]
Sent: Tuesday, August 18, 2009 9:32 AM
To: Bruce Carpenter
Cc: towertalk@contesting.com
Subject: Re: [TowerTalk] Lightning Damage
Bruce Carpenter wrote:
> Les
>
> I have read the message you posted on the lightning protection forum.
>
> First, we must recognize that no level of protection will totally protect
> electronic equipment from lightning damage. We can however reduce the
> probability of damage and the extent of the damage by taking selected
steps.
> Just think of what could have happened if you had not taken the steps you
> did!
>
> It sounds like you have taken some excellent precautions but unfortunately
> you failed to realize the necessity to bond the CATV cable to the ground
> system. I am not sure what is meant by the term SPG.
>
> We must protect each structure individually. EVERY conductive material
that
> enters the structure or maybe an enclosure must be effectively bonded
> together with the shortest and straightest conductors as is practical. It
> is preferable if this is done at a single or nearly a single location.
This
> would generally be at or very close to the electrical service entry point
of
> a structure. The impedance of the bonding conductors used to bond the
> various conductors entering/exiting the structure together is critical as
> significant voltages may be developed between the bonding points depending
> on the length of the bonding conductor. Remember that a 1/4 wavelength or
> odd multiple of a 1/4 wavelength is a high impedance. So, if the bonding
> conductor is 8 feet long and there is lightning energy in the strike at
> about 28 MHz then the bonding conductor is a high impedance and a voltage
of
> several thousand volts may be developed between the ends of this
conductor.
Better not to think of it as a transmission line, rather, a straight
inductance. 8 feet is about 2-3 microhenries (in a straight line). If
the full stroke current flows through it, the voltage can be quite high
(di/dt of 10kA/microsecond * 2uH = 20kV)
> The solution is multiple conductors of various lengths.
That would not be the usual recommended practice. Short and sweet is the
rule. There's really no "resonance" effects one wants to worry about.
The nominal lightning waveform is a so-called double exponential with a
rise time (10% to 90%) of about 2 microseconds and a fall time (to 50%)
of 50 microseconds, which is a broadband signal with a peak around 1 MHz.
Surges on power, phone, cable TV lines are even slower (and actually
cause more damage, typically... direct hits are rare, induced transients
are common)
Remember that
> lightning energy acts like RF, it is conducted on the surface. So, large
> surface area is important. Not braid, but flat copper strap. Braid is
not
> desired because the weaving of the conductors increases the impedance of
the
> conductor and the small conductor size means it is easily broken or
corroded
> and broken.
Flat, round, whatever has surface area. I wouldn't use welding cable
(too many fine strands), but 4/0 and 2/0 stranded copper wire are very
commonly used in lightning protection systems, because they are both
flexible and strong. (you need to allow for the very high mechanical
forces when carrying the lightning transient). Individual lightning air
terminals are often grounded with AWG8 or AWG10, and this sort of size
is what's used for bonding things together (e.g. service panels and
antenns dischage units)
>
> Suitable surge suppression and the bonding of all components together is
> more important than how well the structure is "connected" to the soil or
> earth. The better the connection to the earth then the quicker the charge
> introduced in the structure will be dissipated but those connections must
be
> made at the correct locations. Consider a communications site that is on
a
> rock mountain. There is no effective connection to the soil/earth but
those
> sites generally survive lightning strikes with no or minimal damage. But,
> the protection becomes expensive to some. To others it is cheap insurance
> and just the cost of doing business.
There's always a trade between the cost of transient suppression and
other measures and equipment replacement cost (or insurance for that).
If you're a cell company or public safety agency that needs to be up
24/7 with $1M worth of gear in the shelter, that's a different matter
than a ham with $5K of gear who can tolerate being out of service for a
week. Make sure your house doesn't burn down and nobody will get hurt,
then worry about whether your equipment gets fried. Now, if someone
wants to make a hobby of building lightning protection systems, have at
it, but you have to ask, do you spend $1000 on transient suppression or
$1000 on a new antenna or amplifier or radio?
>
> We also must remember that during a lightning event a significant magnetic
> field is established around the lightning path as well as in conductors
that
> carry the lightning energy to the earth. This magnetic field can be very
> damaging to sensitive electronic equipment that is not suitably protected.
> If you desire to protect equipment from the electromagnetic pulse then all
> of the I/O ports of the equipment must also have suitable surge protection
> devices (SPDs) installed. As an example your alarm system has conductors
> routed to sensors within the structure and external to the structure.
These
> conductors are great antennas! The electromagnetic energy from a nearby
> event is coupled into the wiring throughout the structure and it seeks a
> point where this can be equalized.
>
> You mention you have some SPDs installed. It is important that the
bonding
> conductors be kept as short and straight as possible. The SPDs should be
> tested or replaced once exposed to lightning activity to ensure continued
> protection. You should not use any SPD that is not UL listed for the
> application, excepting maybe RF transmission lines.
Why would you except transmission lines? Either you're following the
code scrupulously, in which case listed components need to be used
everywhere, or you're not. Note that the usual PolyPhaser and ICE
transient suppressors are NOT listed devices.
BUT, a more important comment is that UL listing doesn't mean the device
works better or worse as a transient suppressor, it just means that it
won't start a fire. The much advertised UL1149 is not a "function"
spec, but a set of standard test waveforms to use.
If you need listed devices for code compliance or other regulatory
reasons (e.g. your insurance requires them, or the design engineer
specified them on your construction drawings), then use listed devices.
This mainly applies to
> AC and control line products. Use an SPD suited for the application.
> Consider the operating voltage of the circuit to be protected and select
an
> SPD that has an operating voltage suitably above that point but not so
high
> that the clamping voltage of the SPD exceeds the breakdown voltage of the
> device you are protecting.
Clamping type transient protection (shunt suppressors) can actually
aggravate the damage potential. Consider something like a long telephone
line with a shunt protection device. The telephone line is a
transmission line that has been "charged" by the transient (assuming
we're talking a differential mode transient, not common mode, which the
shunt suppressor wouldn't do anything for anyway). The suppressor
fires, shorting the line. Now, all the stored energy in the line
rapidly flows through the shunt, creating a magnetic field, which
couples into the downstream line, inducing the very voltage you were
trying to suppress.
Series mode transient suppression is the way to go (and it suppresses
common mode too!) or a combination of series and shunt.
Cheap shunt mode suppressors using MOVs also gradually degrade and most
designs do not "fail safe".
Your antenna, rotor control, other
> control/communications, telephone, CCTV and CATV MUST ALL be protected by
a
> suitable SPD. For RF lines you desire a SPD that DC isolates the input
and
> output such as a Polyphaser device. Yes, Polyphaser does make suitable
> devices for CCTV and CATV cables. Many of the inexpensive SPD available
do
> not perform to our expectations. The problem is we don't find out until
> after the fact. Also, we might purchases something at a hamfest that is
> used. Unless we evaluate the product we don't know if it will be
effective
> in handling the surge energy we expect it to handle. Try to avoid used
SPD
> products. Purchase new from quality vendors. The problem is they are not
> inexpensive.
>
> I hope this provides some food for thought.
>
> Regards
> Bruce, W3YVV
>
> Bruce Carpenter Consulting, LLC
> P.O. Box 97
> Dayton, MD 21036-0097
> 410-531-0200
> bruce.carpenter@earthlink.net
> http://www.bruce-carpenter-consulting.com
>
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|