Designing a snubber

[stefanv]

specs for a 4000 watt 60hz snubber?

When the power goes off, I have an in-line inverter system (Trace 2500) t
provide AC to my house. The power lines in my area being above ground
lightning sometimes hits the poles and gives huge spikes and surges
Doesn’t happen often, but when it does the next morning people complai
about there TV’s and computer’s blowing up and so on.
I decided to put a home-brew 4000 watt spike protector before the inverte
to protect whatever is on it. I was thinking some heavy-duty 150V varistor
accros neutral-hot-ground, and a snubber.
About the snubber; looking at some resources, I’m still confused. In m
case, AC power should it be:
- L in series with the hot line, C after that across hot and neutral?
- To slow down the spikes at 60Hz, what should the values be?
- Will this handle 4000 watts?
- Any suggestions on winding my own L (core, turns, wire size)?

Help is appreciated!

Stefan V

 

[Robert Baer]

specs for a 4000 watt 60hz snubber?

When the power goes off, I have an in-line inverter system (Trace 2500) to
provide AC to my house. The power lines in my area being above ground,
lightning sometimes hits the poles and gives huge spikes and surges.
Doesn’t happen often, but when it does the next morning people complain
about there TV’s and computer’s blowing up and so on.
I decided to put a home-brew 4000 watt spike protector before the inverter
to protect whatever is on it. I was thinking some heavy-duty 150V varistors
accros neutral-hot-ground, and a snubber.
About the snubber; looking at some resources, I’m still confused. In my
case, AC power should it be:
- L in series with the hot line, C after that across hot and neutral?
- To slow down the spikes at 60Hz, what should the values be?
- Will this handle 4000 watts?
- Any suggestions on winding my own L (core, turns, wire size)?

Help is appreciated!

Stefan V

Even a cat cannot snub *that* well.

Getting a bit more serious, use a number of lossy ferrite toroids, 3
or 4 in series (glue together to make "fat" toroid) one turn on the hot
side, then a 5-10 turn per line common mode xfmr using same lossy toroid
material; put before anything else.
MOV varistors may be good for fast stuff, but degrade over use - just
like tissue degrades with exposure to radiation.
So they can turn to crap in a short while with a goodly storm.
Use carbide ones - they never degrade.
With something that nasty, i cannot help much.

 

[stefanv]

Even a cat cannot snub *that* well.

Getting a bit more serious, use a number of lossy ferrite toroids, 3
or 4 in series (glue together to make "fat" toroid) one turn on the hot
side, then a 5-10 turn per line common mode xfmr using same lossy toroi

material; put before anything else.
MOV varistors may be good for fast stuff, but degrade over use - jus

like tissue degrades with exposure to radiation.
So they can turn to crap in a short while with a goodly storm.
Use carbide ones - they never degrade.
With something that nasty, i cannot help much.

Thanks for the advice. Not knowing to much about this matter, you wouldn'
put a capacitor at all? Just the toroid in series with the hot? (and som
carbide MOVs after)

Stefan V

 

[Robert Baer]

Thanks for the advice. Not knowing to much about this matter, you wouldn't
put a capacitor at all? Just the toroid in series with the hot? (and some
carbide MOVs after)

Stefan V

A capacitor might help, but with the possibility of rather nasty
spikes, the capacitor would have to be rated to at least 2KV.
The series lossy inductor i think would be best done as one in series
with high side and one in sereis with the low side.
That first tends to phase shift the current WRT voltage as well as
attenuate spikes; next the common mode choke then helps to block (common
mode) spikes.
Get the highest voltage rating common mode choke possible, for 440V
lines or higher...so arcing from one side to the other cannot happen
and compromise the choke.
At that point, a good varistor would seem to be the next step,
followed by another lossy inductor pair (one in series with high side
and one in sereis with the low side), then another common mode choke and
*then* maybe a capacitor (600V or better).
In fact, one could use a spark gap voltage limiter at the start and
across the varistor.
The first one at the front should have the highest current rating
that you can afford (or make it relatively easy to replace and have a
number of them available for replacement after the *big zap*).

Mind you, a lot of what i am "recommending" is guesswork, but i know
that the series inductor pair and common mode choke does work wonders -
from experience.
The use of carbide varistors and spark gap limiters are from from others.

You may need to do a little experimenting, and if game, make a tester
that creates big zaps!

 

[Ian Stirling]

specs for a 4000 watt 60hz snubber?

When the power goes off, I have an in-line inverter system (Trace 2500) to
provide AC to my house. The power lines in my area being above ground,
lightning sometimes hits the poles and gives huge spikes and surges.
Doesn?t happen often, but when it does the next morning people complain
about there TV?s and computer?s blowing up and so on.

4Kw isolation transformer, with a moderate sized low ESR cap on the output?

 

[JosephKK]

Why not?
Are you arguing that the insulation will fail on the primary - which I
have no problem about, or that the transformer will not saturate when
trying to feed this massive pulse to the capacitor (sized to take about
400W capacitively).

An actual lightening strike to the mains will likely explode the transformer
and damage all downstream electronics. More to the point would be
lightening / surge protection at the mains entrance, and again before any
[relatively] sensitive load {like you wish to protect with your
transformer}.