In alt.engineering.electrical Leonard Caillouet wrote:
| wrote in message
| ...
| In alt.tv.tech.hdtv Franc Zabkar wrote:
|
|
| The MOVs will act like conductors when they are clamping. The surge will
| take both paths ... the path through the MOVs, and the path going past the
| MOVs. In general, about 50% will go each way. That can vary at higher
| frequencies.
|
| Why would you assume that 50% will go each way when you don't know the
| impedance of each direction? When conducting, or at failure, the MOV has a
| very low impedance.

There is a distinction between "go each way" and "what comes back" due to
the impedance. It will be about 50% that goes each way _because_ the power
itself does not (yet) know the impedance (at a distance), until it gets
there.

There are two kinds of impedance to deal with here. The first (literally)
is the characteristic impedance. At the point of the MOVs themselves, it
will be about the same each way, but it can vary some at higher frequencies.
It depends on the way the MOVs and the connections with them are constructed.

The second is the net impedance of the path beyond the MOV connections.
That impedance is not what I am talking about in my prior statement. Yes,
it plays a part, but it is not infliienced by the MOVs. It would be the
same if you simply shorted the MOVs with a wire (though that certainly
causes other things to not work, so that isn't how protection is done).

Ultimately you do have to consider the _whole_ system to get an accurate view
of exactly what will happen. Generally this is impractical. What you have
to do is understand what can happen with the variations, and try to change
things to make the happenings do what you prefer (e.g. avoid damage to the
protected devices).

One example involved the power wiring. There should be a point where you
have the neutral grounded, and heavy duty MOVs between each hot wire (be
that 1, 2 or 3) and the grounded wire, and between individual hot wires as
well. The grounded wire (referred to as neutral, but incorrectly in some
cases, even though this is the common referral) would be directly connected
to the path to ground. That connection should be with the least impedance
you can possibly get, within your cost/risk criteria. That means a short
and/or heavy conductor. Short to make it more effective at higher frequencies
by reducing inductive impedance. Heavy to handle a greater current flow.
Much of the surge can now take this path to ground. But not all of it will.
To maximize what will take the path to ground, and minimize what goes to the
building loads/devices that could be damaged, you need to have an increased
impedance on that path. Clearly resistors are not workable since that stops
the power itself, which you do not want to impede. What can work is a low
pass filter primarily an inductor. It needs to be made to have very little
effect at 60 Hz and below, yet block energy/power above that as much as is
possible (again, within your cost/risk criteria). The combination of these
things can limit the surge that reaches protected devices to a tiny fraction
of its original energy.

--
|WARNING: Due to extreme spam, I no longer see any articles originating from |
| Google Groups. If you want your postings to be seen by more readers |
| you will need to find a different place to post on Usenet. |
| Phil Howard KA9WGN (email for humans: first name in lower case at ipal.net) |