"w_tom" wrote in message
...
A surge protector will 'stop' what kilometers of sky could
not?
Indeed it can. The cloud/lightning/cable/earth system forms a voltage
divider: so long as the impedence from cable to ground is significantly
lower than that of the lighning strike, then the voltage will be
(relatively) low.

The main components of the impedence in the divider a
- the resistance of the air to the strike. This is quite low once the arc
is extablished.
- the resistance of any spark gaps to ground. Again, quite low once the arc
is struck.

Power lines are equipped with arcing horns to encourage the ground arc to
strike, so the peak spike voltage once the ground arc has struck is probably
about twice the peak mains supply voltage.

There is a short period between the lightning strike and the ground arc
striking when the voltage on the line can rise to many times the supply
voltage.

Telephone lines don't have this protection, but they do have little green
boxes with very small gaps between the connections inside, which also
encourage arcs, and will certainly have surge suppressors somewhere,
probably at the exchange.

From the strike point to the equipment to be zapped, the main impedence
components a
- The inductance of the cables
- The capacitiance of the cables to ground.
These form a low-pass filter that will tend to spread out the energy pulse,
reducing amplitude and rise/fall time and increasing duration.

Hence, fitting a surge suppressor (only) in a domestic situation will help
if the strike is some distance away - the further the better. It won't help
much is the strike is on the phone wires leading from your local pole to
your house!

If you think that you're at real risk, you can buy commercial units (for
rather more than 69p) which contain a spark gap, a series inductor and a
parallel surge protector. These, if connected to a GOOD earth, will filter
out all but the biggest surges. They are available for power, phone, TV
aerial and data lines.

insult snipped

It is routine to suffer direct strikes without damage even
in regions that have more serious lightning storms.

True, but these systems are designed with lightning strikes in mind.

But the
basic concepts must be understood. Plug-in protectors avoid
all discussion about the most critical function in protection
to sell their ineffective but so profitable products.

These devices are NOT inneffective. The worst you can say is that their
effectiveness is limited. There are limits to all forms of lightning
protection. If lightning actually strikes your TV aerial or phone/mains
incomer, then it is almost impossible to avoid some damage - even if it's
only a zapped surge filter!

Scientists measure direct lightning strikes to a
communication tower atop Hoher Peissenberg in Southern
Germany.

And? What is your point here? Engineers regularly monitor lightning strikes
(and induced surges) on a wide range of systems, including mains supplies,
transmitting & receiving masts, phone systems, water and gas distribution
pipes

Why are those CMOS devices, rated only for tens of
volts, not damaged by million volt lightning strikes?
Most devices in transmitter output stages (and receiver input stages) are,
in fact, rated at hundreds of volts, not tens, but I see the point! They
are not damaged because the designers put in a LOT of effort to fit surge
filtering and suppression between the semiconductors and the exposed bits of
metal.

snipped

Do not confuse the protector with protection. They are
separate devices of a protection system. A protector is only
effective when connected short to protection. That protection
is earth ground.

Active surge protectors are normally connected as shunts (i.e. between the
cable to be protected and ground) and normally present a high resistance as
long as the voltage across them does not exceed a particular threashold,
after which they present a virtual short until the voltage drops back to
almost zero. Hence, they need a series impedence (such as an inductor) to
work repeatedly - if they are connected straight across the mains, they will
short the incoming power to ground as well (for the current half-cycle, at
least) and the current may well overheat them. As they are designed
(usually) to fail short, this will probably blow the incoming supply fuse.
Special devices are needed for telephone systems that have a switch-off
voltage higher than the applied DC, otherwise they stay on for ever once
triggered.

A surge protector is only as effective as its earth ground.

Very true. Earth connections must present not only a low DC resistance to
ground but be capable of carrying more current than the protected cable and
also present a low dynamic impedence. The latter requirement means that the
earth connection should be as short and direct as possible. This tends to
reduce the effectiveness of distribution-board protection where the
protected appliances have an alternative path to ground, due to the
relatively high dynamic impedence of most domestic earth wires. Hence
'protected' TVs, for instance, will often be damaged during a strike on the
aerial by arcs developing between the chassis of the TV and any nearby
earthed metal, such as central heating pipes.

snip

snip I, personally think they are all
a waste of money, but you pays your money etc....

As I said earlier, there is very little that you can do to protect against
direct lightning strikes, but most surges are either induced spikes, where a
strike on a pole causes a large voltage gradient down the pole and across
the nearby earth, and this capacitively couples into the affected cabel, or
mains switching transients. Despite the careful design of substation
switches, you still get arcs when circuits are broken and the surges these
produce can never be completely suppressed at source. These surges have a
much higher source impedence and so can more easily be filtered and
suppressed.

The worst design scenario is the nuclear emp, where you have to protect
against pulses with a risetime of 50kV per millisecond - this edge is so
fast that it often manages to couple itself round standard protection and
fry sensitive electronics, even inside sealed metal boxes...

K