From: John De Armond
Subject: Re: dc to ac power inverters
Date: Sat, 11 Jan 2003 17:58:34 -0500
Well, there are surge protectors and then there are surge protectors. The
el-cheapo ones you buy at Wal-mart and the like contain simple MOVs (metal
oxide varistors) that conduct when the applied voltage exceeds the design
point. The cheapest protectors have one MOV from hot to neutral to clip off
normal mode over voltages which aren't all that common. The slightly better
ones contain 3 MOVs, one across the lines, one from neutral to ground and one
from hot to ground. These will clip both normal and common mode over
voltages. The problem is, most transients contain a lot of high frequency and
the skinny ground wire plus the ground wiring back to earth ground present a
high impedance to the high frequency components. Thus, while the MOVs may
conduct, there is no place for the transient energy to go to.
The best surge suppressor contain staged protection. It is assumed (good one)
that no one stage can handle the range of events possible. So there are
several different devices, one after the other, each of which addresses one
form of disturbance and which dissipates part of a large one.
Generally the first stage is a set of gas discharge tubes shunted by
capacitors. The gas discharge tube works like the MOV - conducting when a
threshold voltage is reached - but it can handle and dissipate much more
energy than an MOV. The caps shunt off the high frequency stuff that has a
rise time too fast for the gas tube to fire in time. Following that may be a
series of MOVs like I described above, perhaps separated from the gas tubes by
a longitudinal choke. The idea is the MOVs will clip what the gas tube misses
and the choke will provide some impedance to limit the peak current.
Following that will be an LC low pass filter. This is a block that contains a
series of inductors and capacitors and is designed to block all energy higher
in frequency than 60 hz. That takes care of the energy that gets through the
MOVs as well as harmonics that won't trigger the MOVs but WILL cause some iron
cored inductive devices (motors, transformers) to over heat. Last is another
set of MOVs, these with a voltage setpoint lower than the ones on the other
side. They'll clip off any high voltage ringing that may escape the LC
Needless to say, such a suppressor can't be had for $4.95.
The problem with this high quality suppressor relative to inverters is that
both the caps that shunt the gas tubes and the caps that constitute the input
stage(s) of the LC filter present a momentary short to the inverter's output
FETs. The problem is similar to capacitor input battery chargers that I
described in detail in another post.
Since the capacitors are smaller than in a battery charger, the short circuit
current may not be high enough to trigger the inverter's protection circuitry.
It may, however, cause the suppressor and/or the inverter to overheat, output
funky waveforms and draw high idling current.
There is no need for a suppressor on the output of an inverter. The el-cheapo
"surge strips" that contain nothing more than MOVs won't hurt anything.
Usually. Unless the mfr happened to throw some moderately high value
capacitors across the MOVs.
I use the cheap strips with no problems on my inverters. I have also
experienced all the described problems trying to use a good suppressor on the
While I'm on the subject I should note that no single suppressor, regardless
of the quality, can provide protection against large surges, particularly
those that come riding in on all three wires. Such as lightning-induced ones.
The house wiring leading back to earth ground is simply too high an impedance
to conduct it away. Especially when the surge is on the green wire itself.
The solution is to provide distributed protection with the design philosophy
being to deflect/dissipate the surge as close to the service entrance as
possible. This means an arc gap and/or gas tube suppressor in the meter base
or even better, up at the peckerhead, another gas tube/MOV combo in the
breaker panel, perhaps filtered and MOVed outlets and finally a good
suppressor at the load.
All devices are available, though not at Home depot and the like. HD's "whole
house suppressor" that mounts in the breaker panel is pretty good - I've
tested it and have found it works as advertised. There are better ones but
they cost more money. Lightning Protection Associates Inc makes a line of
protection equipment designed to protect against the worst environments -
Florida and mountain top radio sites.
Several utilities including mine, are now offering meter box suppressor.
These install under the meter and work very well. Unfortunately most
utilities are scamming people into monthly rentals instead of just selling
them. Locally they get $5 a month for life for an $80 unit.
Back to RVing, If I used any line-operated devices that were surge sensitive
in my rig, I'd use a high quality suppressor on the line cord near the plug
and a whole house suppressor on the breaker panel. I'd also lay my power cord
out with a loop or two in it between the pedestal and the RV, the loops
providing enough inductance to shunt off an amazing amount of lightning surge
energy. NOT grounding the RV (other than through the green wire) helps too.
On Sat, 11 Jan 2003 07:43:39 -0500, Barrie Brozenske <firstname.lastname@example.org>
>In article <3E1F877C.85ECF41@hotmail.com>, email@example.com says...
>> And do not plug a power
>> strip surge protector in the output of a converter - be it the
>> vehicle one or the one inside a plug-in UPS.
>Tom, I take it you mean "do not plug a surge protector into the output of
>an INVERTER"? Why not? I have mine plugged into my triplite inverter,
>works fine. I mainly did it to get more outlets.
From: John De Armond
Subject: Re: zero-surge or isobar tripp-lite?
Date: Mon, 30 Jul 2007 17:08:32 -0400
On Mon, 30 Jul 2007 11:15:09 -0700, bud-- <firstname.lastname@example.org> wrote:
>> Where does APC cite each type of surge and claim to protect from
>> that surge? Why do they not make that claim?
>Absolute nonsense. Plug-in suppressors have MOVs from H-G, N-G, H-N.
>That is all possible combinations and all possible surge modes.
Actually those only scratch the surface.
Common mode induced surge on the power line
Common mode induced surge on the phone line
Common mode induced surge on the cable TV line
Common mode induced surge on the TV, radio or ham radio antenna line.
Common mode surges from the power system to something else better grounded. Ex:
Normal mode surges from things such as switching transients aren't all that common in
residential settings, contrary to propaganda from surge suppressor marketers. This
statement is based on years of power quality engineering consulting and to the
Dranetz Line Disturbance Analyzer that operates 24/7 in my shop.
Much more common is the common mode surge where all conductors elevate above earth
reference. Usually caused by lightning in the vicinity, these are quite damaging
because they cause breakdowns between the device's works and surrounding conductive
surfaces which are not affected by the transient. Arcing from the ground plane of a
PCB to the metal case is commonly seen.
Outlet strip surge suppressors do nothing against this type of surge simply because
there is no potential difference between the conductors in the gadget. The only way
to stop external surges is with suppressors tightly coupled to low impedance true
earth ground. In practical terms that means meter base suppressors and whole house
If the impedance to earth ground is low (including for high frequencies
characteristic of lightning-induced surges) then the suppressor will conduct away
surges originating both outside the house and those induced in the house wiring, as
long as the wire length isn't too long.
Outlet strip surge suppressors have their place, of course, but to believe that
plugging appliances into one of those strips addresses "all possible combinations" is
From: John De Armond
Subject: Re: zero-surge or isobar tripp-lite?
Date: Fri, 27 Jul 2007 14:32:20 -0400
On Fri, 27 Jul 2007 11:09:40 -0400, email@example.com wrote:
>The best protection (if you remember to do it) is to plug all
>components connected to a computer into a single strip, and then
>physically disconnect it during a storm, along with any fax/modem or
And even then sometimes that isn't enough. Lightning struck a tree perhaps 200 ft
away awhile back. Nothing connected to the power or phone lines was damaged even
though I heard the arc and saw the little fire ball roll out the surge-suppressed
outlet. The portable USB hard drive that was disconnected and stored in my computer
bag wasn't so lucky. Tango uniform. Apparently the USB cable was enough of an
antenna to pick up enough surge to kill the thing.
>I had an interesting example of lightning quirkiness the other day. A
>neighbor had a significant strike on a poplar tree - enough to blow
>poplar bark a good 50 feet onto the roof of his house, and leave an
>open split on the tree. One of my computers was connected to the
>phone line at the time, but totally disconnected from the power lines
>and sitting 500 feet away in a dry house with a metal roof, where
>leakage to ground was unlikely. The modem in the computer blew.
>Sometimes stuff happens. (Yeah, I know, I forgot to unplug the modem.)
Probably not a stray leader, though. More likely a simple inductive spike induced in
the phone wiring from the high amperage of the lightning bolt.
>In networked situations with hard wiring, I've had very good luck with
>the Isobars and Isotels, even with direct strikes on the building, and
>in one case a strike that took out the main electrical entrance, and
>would have fried any surge suppressor that attempted to cover the
I like Isobars too but even they can't do the job by themselves. A defense-in-depth
approach whereby the surge energy is redirected to ground at several points and
attenuated at each point is the best one. I've quite a bit of experience keeping
radio electronic alive on top of mountains and this is the approach I take.
My cabin here is quite similar to a mountain-top radio installation. I'm at the end
of a 30 mile feeder, fairly high on the mountain. My Dranetz power quality analyzer
sometimes can't print fast enough....
I have a meter base surge arrester, a breaker panel arrester and arresters on each
outlet. The phone line is looped several times to provide inductive impedance to the
surge and there is a spark gap upstream of the coil to divert the energy to ground.
There is the usual surge arrester at the demarc and another inside at the fan-out
panel for my communications wiring. All the field wiring for the burglar alarm
enters the panel through a phone-type surge arrester punch-down block.
My cabin is ringed in buried #4 copper cable with multiple ground rods. The soil
here is a sand-clay mix and is always wet so ground conductivity is great, some of
the best I've ever tested.
Knock on wood, in the 30+ years I've had the place I've never lost a single
electrical or communications gadget. That hard drive was the first and it wasn't
even connected to anything.