> the ONLY function of a lightning arrester is to equalize the voltage between
> conductors. It clamps the voltage between conductors to some low value that
> is designed to prevent damage to other components. They are really only
> effective if they are close to the equipment being protected since any
> exposed conductor may allow induced voltages to accumulate. And they have
> to be on ALL power and signal conductors at the same location.
And there's another subtle problem that crops up with these gas tube or
fast diode type suppressors (anything that breaks down and has a very
low impedance after breakdown)..
And that's the fact that a gas tube can turn a slow transient into a
fast one (just like in a line pulser). Two scenarios are possible...
Big loop outside the suppressor: a worst case might be something like a
phone line coming in from the pole, and the suppressor grounding to low
impedance at the entry point. The voltage builds up on the phone line
(induced from a lightning nearby, or striking the line somewhere a
distance away).. the gas tube fires, and now, you basically discharge
the stored charge in the drop. That fast discharge creates a magnetic
field which induces a voltage on any loops *inside* the protected area.
(relatively few of us run our radios in a magnetically shielded room)...
Because the magnitude of the field transient is proportional to di/dt, a
fast tube makes it worse!
Second scenario.. shunt mode suppressor is connected across a pair of
wires. Voltage rises on pair, gas tube fires, shorting the pair.
Traveling step wave propagates both directions from the gas tube.
You've turned a slow rising transient into a fast edge pulse. Depending
on what's downstream, you may have produced something that capacitively
couples into the victim equipment. Or, the transient that is
propagating back towards the source hits another discontinuity (another
gas tube that has fired?) reflects back, and lines up on top of other
This is a real big problem with power lines and switching transients..
voltage rises of 3-4 times are not unusual if the timing is "just
wrong".. which is why they go through a fairly careful process of
"coordination", making sure the breakdown voltage of protective gaps is
appropriate at each place in the distribution system. When lightning
strikes the power line, you want it shunted somewhere close, but you
really don't want to make things worse with a cascading transient trip.
(and it doesn't always work right... )
TowerTalk mailing list