[Top] [All Lists]

Re: [TowerTalk] whole house surge protectors

To: Jim Lux <>,Tower Talk List <>
Subject: Re: [TowerTalk] whole house surge protectors
From: John Elsik <>
Date: Wed, 19 Sep 2007 14:32:45 -0600
List-post: <>
I have been reading this with great interest.  I am installing a ground system 
now.  I was going to use an I.C.E 330 for the AC mains (whole house).  Would 
that be adequate?  Something is better than nothing.  It looks like it is a 
shunt type also.
John wa5zup

> Date: Wed, 19 Sep 2007 10:34:55 -0700> From:> To: 
>> Subject: [TowerTalk] whole house surge protectors> 
> > I've been doing a bit of browsing> Most of the whole house protectors are 
> shunt mode (which, by the way, > can actually cause some problems, making 
> things worse)> > Intermatic IG1240RC> Leviton 51120-1> Panamax gpp8005> 
> Siemens various models> > etc.> > In any case, a shunt mode suppressor will 
> also suppress transients > originating within the house.> > There is a case 
> where they will be of less effectiveness.. if you have a > transient induced 
> on the branch circuit between the panel and your load, > then the transient 
> propagates both directions, and, depending on the > relative lengths of the 
> wire, it will get to the load before it gets to > the clamp. Once the clamp 
> goes into effect, an inverted transient gets > reflected back, so you can 
> calculate the maximum width of the > unsuppressed pulse. (you could use 
> something like 
 2 ns/ft as a > propagation speed... so for a 100 ft run, with the transient 
induced > next to the load (or downstream from the load), you get half a > 
microsecond or so before the voltage is clamped.> > > If the transient is 
induced on a branch circuit other than the one your > load is on, the 
suppressor is between the transient source and your > load, so it would clamp 
the transient before it arrives at the load.> > If you put a point of use 
transient suppressor *with a higher voltage* > than the whole house protector, 
it will take care of the half > microsecond impulse before the whole house 
protector kicks in, and won't > have to absorb as much energy.> > FWIW, 
statistics show that most transient damage occurs from transients > originating 
outside the house, typically from a lightning strike or MV > line / LV line 
fault somewhere (MV = 10-30kV, LV = 120,240,480V). That > makes the rise time 
of the transient much slower (it's low pass filtered > by the power line), and 
 o makes the whole house protector more > effective.> > In the event of a MV/LV 
short (the only kind of line transient I've had > personal experience with in 
Southern California.. we don't have much > lightning here), you've probably got 
a significant overvoltage that > lasts 8-10 milliseconds or longer (until the 
MV breaker trips or fuse > blows). You'd have to hope that the surge suppressor 
can hold on that > long without blowing its internal fuses. Or, ideally, it 
would short, > and trip your main breaker. Since most of these whole house 
protectors > have energy absorptions in the few kilojoule range, I'm not very > 
sanguine about their ability to protect against this kind of fault. > Figure 
the case of a 14.4kV feeder shorting to the 240V drop into your > house (this 
has actually happened to me). If the series impedance of > the feeder, through 
the drop, into the grounding system, is on the order > of an ohm or so, the 
fault current is around 10-20kA.> > If the surge prot
 ector is based on an MOV, the 400V MOV is going to > dissipate about 4-8 MW, 
or, 4-8 kJ/millisecond. If it takes a half > cycle for the breaker to trip/fuse 
blow, that's 8 ms, and about 50kJ > (which is why things literally explode when 
this happens).> > A surge protector based on a spark gap, which, once it fires, 
has a much > lower clamping voltage, will dissipate less energy in the 
protector, so > has better survivability.> > 
_______________________________________________> > > > 
_______________________________________________> TowerTalk mailing list>>

TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>