Re: Re: Re: Optoisolation for a modem

From: "Jim Fackert" <jfackert@...>
Subject: Re: Re: Re: Optoisolation for a modem

So the impedance of the resistors is adequate, with the strike shorted
to
ground, to limit the voltage and protect the parts, adn you DON'T want
them
to open up like fuses! My mistake. Instead of the hard to find carbon
comp, you could also use a 2 to 5 watt wire wound... the inductance
might
have some (positive?) effect on the surge suppression though...

Wirewounds have a tendency to arc over also, at least the older ones.
The two good behaviors are in order of preference are:
1. Drop the incoming spike and survive.
2. Open up and block the incoming spike. (fuse)

Any arc overs tend to negate the impedance/resistance of the circuit,
very undesirable.

From: George Potter <gpotter@...>

>A couple of notes ......

>The FCC part governing phone line devices is Part 68 ...... and the "lightning

tests" required therin DON'T require the equipment to survive a 1500
volt
strike .....

Correct, however a subscriber in parts of Colorado and other areas of
heavy thunderstorm activity won't be happy about such a wimpy interface.

>In the US, much of the lightning "protection" (on the phone company side) is

via "carbon blocks" that will conduct transients to ground (to some
degree).

Colorado was using gas tubes, but your are could be using this
technique. George has a good point, make sure you have gas tubes and
that they are fairly new, less than 5-10 years old.

>I personally like gas tubes, because they do react very fast, and survive many

strikes, however their downside is their firing voltage (being somewhat
higher
than many MOVs).

That's why we used both.

>The UK phone intefaces I designed didn't have any lightning protection, as

they (by design and component selection) would survive 5 KV dc line to
ground strikes without failure). (and at that time, there wasn't any
surge
requirements, but because of the various "barrier" requirements we ended
up with a tough interface).

That is very tough! Anything special up front?

From: Jon Elson <jmelson@...>

>Well, I can say that my setup, with the film resistors, has survived

about 4 years in St. Louis, and
the 'Network Interface Unit' on the outside wall of the house has had
it's supressor blown out
at least once since I put mine in. (Of course, the Phone company's unit
gets to handle as much
of the surge as it can, FIRST!)

Sounds like the newer film resistors are pretty tough, as long as they
open or survive, their OK! St. Louis gets more than it's fair share of
electrical activity.

From: Jon Elson <jmelson@...>

>No, I don't think a 10 Ohm resistor is enough to limit the current of a

1000 V or higher spike! I DO want them to open up, like fuses, and
quickly! I'd use a higher value, but it will start to affect the line
characteristics.

We tested it with simulated strikes and field trialed it. We had about
12 units survive behind blown out demarcs also in addition to just
stopping the wimpier strikes. 10 Ohms is the best compromise, also
recommended by others on the list. No units failed due to lightning that
we could tell.

It is your own house, make sure you you have gas tubes, then some kind
of well layed out MOV device built using any of the devices recommended
and maybe an additional unit inside your house of some kind if your
modem is especially tender.

Good luck,

Jim Eckman