Towertalk
[Top] [All Lists]

Re: [TowerTalk] Whole House Surge Protectors

To: David Jordan <wa3gin@erols.com>,Kelly Johnson <n6kj.kelly@gmail.com>
Subject: Re: [TowerTalk] Whole House Surge Protectors
From: Jim Lux <jimlux@earthlink.net>
Date: Tue, 10 Apr 2007 21:42:59 -0700
List-post: <mailto:towertalk@contesting.com>
At 12:10 PM 4/10/2007, David Jordan wrote:
>Cutler Hammer...great little unit that bolts on to your breaker panel.
>Remember that you'll need end point units at all your sensitive
>electronic equipment outlets as well.

Actually, not entirely true...  If you have a decent whole house 
system, then the point of use supression should be chosen very 
carefully.. for instance, you want the threshold voltage to be HIGHER 
than the threshold voltage for the whole house unit.  If you could 
find a good series mode suppressor (essentially a CLC or LC network) 
that would be good.  What you definitely do NOT want is something 
with a threshold voltage LOWER than the voltage of the suppressor at 
the entrance panel. The challenge will be finding something, since 
the run of the mill plug strip thingees are sold by "clipping 
voltage" where lower is better.

(Unless you have something inductively coupling into your house wiring....)

I believe that K9YC has a description of this on his website, and 
Ronald Standler  does too.

>The whole house units get the big
>stuff and are designed to self restore for numerous hits over years of
>time...to do that they have to limit the clamping performed, so some
>energy is allowed to pass and that is where the end point units
>pick-up.  They are more sensitive and handle the last bits of energy
>that may threaten your equipment.

Extensive testing of modern consumer equipment has shown that they 
can tolerate kilovolt sized impulses on the AC line input. A typical 
whole house device might clamp at 220 to 250 V or so.  The whole 
clamping voltage thing is driven by marketing, leading to horrible 
things like 170V MOVs on 115V power lines, where if the line voltage 
is even a little bit high, the MOV is essentially conducting on every 
half cycle, leading to premature failure, fires, etc.

There are very few electronic devices that can't take a few 
microsecond pulse at 250V (especially all those widgets with 
switching power supplies... they've got diodes feeding a big bulk capacitor)



>For my really expensive gear, ham radio bench and HDTV I use a
>Un-interruptible Power Supply that runs off internal battery all the
>time and therefore hasn't got any switching gliches and spikes on the
>load side like the less expensive UPS systems.  If you are thinking
>about using a UPS be sure you pay the extra bucks to get the FULL TIME
>UPS rather than the transfer switch type.  Both unit types have
>extensive surge protection, so the money you were going to put into end
>point surge protection can be subtracted from the cost of the UPS.

Or, just buy decent CLC filters.. the lightning impulses, as 
manifested on your power line, are fairly slow (tens of microsecond 
rise time), so you need a LOT of L and C.  Fast high voltage 
transient turns into slow, lower voltage transient, and the power 
supply in your device just rides through.

A static inverter UPS might be overkill (and will certainly be a 
power consumer.. overall efficiency might be 80% or worse..)

In any case, shunt mode suppressors have a lot of unexpected side effects.

A lot has been learned about actual transients and device 
characteristics in the last 20 years. 20 years ago, it was tough to 
even accurately record transients in the field, so there were some 
estimates of what might be the case, based on very few actual 
measurements. Power supply and appliance design has changed a lot in 
20 years too, as well as the tolerance for transients. Analytical and 
modeling techniques have also improved.. today, it's feasible to 
actually model the transients on a power line and to model the 
behavior of the nonlinear device.  In 1985, you'd have been lucky if 
you had a version of SPICE and some handbuilt models of basically 
linear devices and some time on a mainframe or a 8088 based PC to run it on.

What might have been a state of the art transient suppression 
approach in 1990 is likely not considered optimum today.  Even the 
components being used to do the suppression have changed.

Just because shunt mode MOV type transient protectors are still sold 
doesn't mean that they are a good solution (they might be, they might 
not.. depends on the application). That's determined more by whether 
the company that sells them can still make a profit.

Before spending significant (>$50) sums of money on transient 
suppression, it's worth it to spend the $20 on books like Standler's 
to find out the real deal..It is truly an eye-opener...I've 
completely changed my thinking about what's actually needed and 
what's appropriate, especially after going and looking up some of the 
papers and reports he references.  And, for what it's worth, a lot of 
the recommendations are embodied in the latest IEEE Emerald Book 
(IEEE Std-1100), which is sort of the bible for this sort of 
thing.  The Emerald Book is a great reference, but quite pricey at 
$140 or so...If you've got significant investment in equipment and 
are really interested in "doing it right" that's the tome to get (or 
borrow.. you can probably get it from your local library on 
interlibrary loan).. Make sure you get the 2005 edition..There are 
signficant changes from the 1999 edition and even more from the 1992 edition.


Jim, W6RMK  


_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

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