Power Surge

[Vanguard]

As regards for *my* personal checklist that says to get a UPS with a
sine wave output, yes, it is unnecessary if all you run from that UPS is
your computer. I know the computer power supply will handle stepped
wave output. Probably the printer, monitor, scanner, speakers, and
modem, too. But eventually you upgrade your system and want a bigger
UPS. In my case, I want a quieter one; the 2kVA unit that I have now
has a fan that makes anywhere from 35 to 50 dB of noise. Will I toss it
because I want one without a fan? No. Will I toss it when the
batteries go bad? No, because they are replaceable; I've already
replaced them once after 5-1/2 years of use (although I decided to use a
bus bar instead of a heavy wire lead) at a cost of $140 for the 2 of
them which was a tenth of buying a new unit. Since I'll use it for
something else, like my audio/video gear, having a sine wave output is
important. It gives me more choices on how to use the UPS other than
just with computer gear.

And for those of you that like pictures with the explanation:

http://www.explan.co.uk/antisurge/modes.shtml

But apparently it also depends on which country you live and their
wiring standards:

http://www.zerosurge.com/HTML/mode2.html
http://www.zerosurge.com/HTML/works.html
http://www.zerosurge.com/HTML/versus.html

In the U.S., the neutral is bonded to earth ground at the service
entrance, the only way to have common mode surges is from something
inside your house, not from outside. Wouldn't a common mode surge
coming at you from outside become a normal mode surge (since it then
proceeds only on the [hot] line)? I think all the surge arrestors that
I've seen here are mode 1 only (line-to-neutral) because neutral was
tied to local ground (earth) at the service entry.

No, I don't work for ZeroSurge and didn't know about their products
until this thread had me hunting around on surge protection and methods.
Their article at http://www.zerosurge.com/HTML/about.html had me
grinning because I remember calling the telephone company to report
problems. Three times they came out and did their local ring test. No
problems. The fourth time, after looking around myself, I stood there
while they tested. No problems again using both their test gear and
making test calls. Took them to the basement and pointed at the MOV
(metal oxide varistor) used across their lines. Replaced it, problem
fixed. The problems I experienced were indicative of a high resistance
short and I figured the MOV went bad. The MOV has very high resistance
(10^9 ohms) at low voltages and passes very little current, but it has
very low resistance at high voltages which shunts the surge but only for
short-duration impulses. That's why they are rated in Joules (N joules
= N amps across 1 ohm for *1 second*). MOVs can burn out with one good
surge.

Most surge protectors rely on MOVs but these eventually go bad over time
after being repeatedly stressed. Several grains of zinc oxide (each
with an individual breakdown of 3 volts) are sintered together to make a
MOV. Each transient destroys some of those grains which reduces it's
level of protection. The more it has protected you, the less it will
protect you. The eventual failure is a short circuit (which is why
sometimes they'll blow apart) and why a fuse should be used. If not
fused, you'll smell when a MOV eventually shorts; it will smell like an
electrical fire. If MOVs are not themselves fused or enclosed in a fire
retardant, they can start fires (see http://www.smarthome.com/4401.html
for a news report). Get a surge protector that has an LED for status to
tell if the MOV is still good. The MOV should be [thermally] fuse
protected or uses disconnect circuitry to shutoff the surge protector.
That surge protector becomes nothing more than an extension cord when
(not if) the MOV goes bad. A surge arrestor at the mains that goes bad
is worthless, too. I can see how an LED can indicate if a TMOV is still
good (http://snurl.com/2824), presuming the thermal element fries before
the MOV has overly degraded. I suppose for a MOV you could do a similar
setup in using a fuse and tie the LED between them.

I've seen users employ surge protectors and UPS'es to protect the
computer and forget about the telephone line to their modem or the cable
going to their cable modem, especially if an internal modem, or even to
their expensive audio gear. Also, to those that extol installing a
whole-home surge arrestor (which may incur the expense of an electrician
and requires it be your house to toy with), a whole-home surge arrestor
is a point-of-entry device suppressing surges stemming from outside
sources (utility company problems, transformer switching, etc.) but it
won't do anything to suppress the high number of power surges that
originate *inside* your house (appliances, projector bulb burnout,
motors, air conditioning, etc.). I just saw some white papers regarding
the origin of surges but had a problem in IE, had to close a window, and
lost all of them. My recollection was that 20% of surges come from
outside and 80% come from inside. Your A/C causing a surge is not
usually a problem for your computer's power supply, but that huge lathe
in an industrial environment is.

While I can see the arguments for using a point-of-entry or
service-entry surge arrestor for the whole home (if that is really an
option for you, like you own that house), that may not eliminate surges
generated within the home so I still see value for point-of-use surge
protectors. w_tom extols dumping the surge to ground but neglects that
data lines and audio equipment use this as their reference so the surges
that are getting pummeled into local common can cause data and audio
problems. It's a 20-year old spec based on using stand-alone devices
and shunting the surges. Surge voltage on the ground wire varies by its
length. With all the interconnected devices in computer systems, and if
you tie in your audio/video system, too, you will have ground loops. If
the surge induced on the safety ground at one computer is 1000V but it
is 600V at another computer or printer, the 400V difference can cause
disruption, degradation, or damage. Even small surges pumped into the
safety ground wire can cause data errors, lockups, or other "weird"
problems.

 

[w_tom]

Common mode surge does not become differential once inside
house. It does not because voltage difference remains
between wire with surge and earth ground. That common mode
surge inside the building is still seeking earth ground,
destructively, via appliances.

However common mode surges can induce differential mode
transients. Those long wires bundled with other wires will
induce transients onto those other wires. This is the induced
surge - not from electromagnetic fields of a distant CG
strike. A plug-in protector tries to earth the common mode
surge down safety ground wire. But that long ground wire,
bundled with other wires, only induces the transient on those
other wires. Just another reason why plug-in protectors are
not effective as noted by last paragraph in:
http://www.explan.co.uk/antisurge/modes.shtml

Zerosurge and its competition Brickwall and Surgex are
series mode protectors. Basically just low pass filters.
They cannot work like a dam. Surges will simply overwhelm the
dam. But they do work well as dikes. Difference: a dam must
stop the flow whereas a dike only diverts the flow
downstream. For series mode protectors to work, the surge
must have a downstream path to earth ground. Series mode
protectors will enhance the protection AND should also reduce
noise on that circuit. But, they still require the surge to
be provided a path to earth ground - and that is what shunt
mode protectors do. Shunt mode protectors make a connection
to earth ground only during the surge.

Second reason why series mode protector require that shunt
mode earth ground: the safety ground wire completely bypasses
the filter. If surge is not earthed at service entrance, then
surge will get to electronics by safety ground wire,
unimpeded, right through series mode protector. They forget
to mention that part in the Zerosurge summary. Even series
mode protectors have this requirement - a surge protector is
only as effective as its earth ground.

MOVs do wear out with use. They are useless for small,
daily transients such as spikes from a turned off switch. But
then those spikes are so trivial as to only be called noise.
So small that MOVs don't even see the transients. If those
daily spikes caused damage, then the homeowner was replacing
dimmer switches daily. For audio recording equipment, a
series mode protector would keep those noise spikes away from
recording equipment. But that noise from switches and motors
does not damage electronics. If noise did damage, then we
were all replacing clock radios every day - long before the PC
even existed.

MOV life expectancy is determined by joules. Since most
plug-in protectors do not even claim to provide effective
protection, they are also undersized - too few joules. As
joules increase, the life expectancy of a surge protector
increases exponentially. For example if a 345 joule protector
(only using as little as 115 joules in protection) were to
survive three same size, standard, 8/20 usec surges, then the
1000 joule 'whole house' protector would withstand on the
order of 400 same sized transients. Numbers that plug-in
manufacturers would rather avoid.

Since lightning surges occur typically once every eight
years, then a properly sized 'whole house' protector using
MOVs will last more than long enough. They just forget to
apply numbers when hyping MOV fear in that Zerosurge
discussion.

If using a 'whole house' protector, then any surge, interior
or exterior, is earthed before it can create damage. Nothing
inside the house should create a common mode surge. But even
if something does, the 'whole house' protector still shunts
the common mode transient to earth.

Some papers will outrightly confuse. For example, 80% of
transients can come from internal sources. IOW a 5 volt noise
spike can be generated by an electric motor. If paper is
written by an English major, then that 5 volts suddenly
transpires into a 500 volt surge! In reality, if anything is
creating destructive surges inside a house, then the homeowner
is replacing dimmer switches, clock radios, and GFCI (RCD)
daily. Appliances only create noise - well below what every
household appliance must withstand. Surges come from exterior
events. But too many authors on surges have eyes that glaze
over as soon as numbers are provided. These people see surge
everywhere - which is what plug-in manufacturers live off -
misinformation.

But lets assume that central air conditioner is generating
surges. OK. The surge will immediately take out internal
controls, furnace controls, and the house electronic
thermostat. But if the air conditioner is creating this
surge, then the surge protector must be at the air conditioner
- not on everything else inside the building.

And since central air is a dedicated circuit, then the
'whole house' protector would eliminate that mythical surge.
Air conditioners and other household appliances don't create
surges. They create noise - which hypsters of plug-in
protectors forget to mention since they don't provide (or
learn) numbers.

Want to see an example of outright lies because the author
probably did not understand fundamental concepts and probably
feared numbers?
http://www.howstuffworks.com/surge-protector.htm
Facts so perverted as to be one long chain of lies. Many who
promote surge protector don't even have basic electrical
knowledge such as this author.

Important for the surge protector to be at the service
entrance for exterior common mode surge since those are higher
frequency events. That high frequency in lightning generated
surges are why the connection to earth ground must be less
than 3 meters. At higher frequencies, wire length takes on a
whole different characteristic. It is why a line - neutral
protector is effective when the connection of line to neutral
to earth ground rod is less than 3 meters (and other
electrical characteristics such as not inside metallic
conduit, no sharp bends, etc). Wire length is critical to
where a surge protector is even connected to earth ground.
Earth ground too far away all but does not exist - because of
electrical characteristics of destructive surges.

Your MOV on phone line is why MOVs are no longer used for
such protectors. Semiconductors have long since replaced
those MOVs. They either work, or become short circuits that
tell human to replace them. MOVs have too much capacitance
for use on phone lines. However some pathetic plug-in
protectors do use MOVs for phone line protection. Your
current phone line protector should be semiconductor based as
has been standard for about 15 years. And it should be each
phone wire, through protector, less than 3 meters to single
point ground. Wire to wire protector do comparatively little.

LED to indicate a surge protector as good is really
nonsense. That LED can report one mode of failure, but cannot
report the protector as good. This example from Zerosurge
demonstrates the problem. They even removed all MOVs and the
indicator light still declared the protector as OK:
http://www.zerosurge.com/HTML/movs.html


Transients on internal safety ground wires are irrelevant -
basically don't exist - if the 'whole house' protector is
installed on every incoming utility (or that utility is
directly earthed before entering the building). If using
internal plug-in protectors, then transients would be induced
every where on safety grounds. Described above were induced
surges. But then ground loops would use nearby appliances as
a path to earth ground because breaker box earth ground is too
far away. This is why an adjacent plug-in protector on AC
input to a computer can contribute to damage of a modem - even
if computer is powered off. That is correct. Adjacent
plug-in surge protector completes a circuit to earth ground,
destructively via modern. And modems, especially in North
America, are most often damaged by AC electric surges.

Once a surge has gotten inside the building, then it will
find numerous destructive paths. No plug-in protector is
going to stop that. In fact plug-in protectors can give that
surge more potentially destructive paths to earth ground -
such as the above modem example.

Can MOVs create a fire? That is what UL1449 2nd Edition is
about. Don't know of any ISO equivalent. But then if those
'adjacent to appliance' MOVs were effective, then MOVs are
already inside the appliance. Appliances already have
effective internal protection. Protection that assumes the
'big' transient - the surge - will be earthed before it can
enter the building. Internal protection that makes adjacent
plug-in protectors redundant. Internal protection that makes
any other appliance generated 'noise' (not surge) irrelevant.

Do you need an electrician? Even if plug-in protectors are
installed, the electrician would still be necessary to upgrade
a house ground. Many homeowners can install these 'whole
house' protectors themselves - since it is that easy - and
therefore sold in Home Dept. Even plug-in protectors require
earth grounds. Older homes are too often missing this
connection that is also essential to human safety. In North
America, earth ground should be upgraded to, minimally, post
1990 NEC code requirements. Equivalent earthing is necessary
throughout the world. That AC electric earth ground must be
less than 10 feet from main disconnect (where neutral meets
earth ground) and that earth ground (sometimes called GEC).
And all other incoming utilities must also connect to that
same earth ground.

Exampled was a ground loop problem of computer at 1000 volts
and printer at 400 volts. These problems are created when
single point grounds are not used. Single point grounds that
are necessary and provided by 'whole house' protectors.
Resulting voltage difference is called ground loops. It is
why peripherals connected to a computer should share same
receptacle outlet if not same circuit (or ring). These
transient voltages occur when the surge is not first earthed
at the service entrance. These voltages are not generated by
internal equipment. If any household device generates such
voltages, then either a surge protector must be permanently
connected to that appliance (the absurd solution) or the
device must be removed as a threat to human safety. But then
that device would remove itself; would self destruct anyway.

Devices have more than sufficient internal protection even
on interconnected ports IF the major surges are earthed before
entering the building. Every incoming utility must either
make a direct connection to single point earth ground OR make
that connection via a 'whole house' protector. Plug-in
protectors only provide surges with more destructive paths to
earth ground such as via appliances. Plug-in protectors sell
by forgetting to mention critical earthing OR provide any
useful numbers. Plug-in protectors want you to not even
discuss common mode transients because they are not selling
effective protection. They do not even claim to provide
common mode surge protection.

One fact always remains - a surge protector is only as
effective as its earth ground. No way around that fact.

As regards for *my* personal checklist that says to get a UPS with a
sine wave output, yes, it is unnecessary if all you run from that UPS is
your computer. I know the computer power supply will handle stepped
wave output. Probably the printer, monitor, scanner, speakers, and
modem, too. But eventually you upgrade your system and want a bigger
UPS. In my case, I want a quieter one; the 2kVA unit that I have now
has a fan that makes anywhere from 35 to 50 dB of noise. Will I toss it
because I want one without a fan? No. Will I toss it when the
batteries go bad? No, because they are replaceable; I've already
replaced them once after 5-1/2 years of use (although I decided to use a
bus bar instead of a heavy wire lead) at a cost of $140 for the 2 of
them which was a tenth of buying a new unit. Since I'll use it for
something else, like my audio/video gear, having a sine wave output is
important. It gives me more choices on how to use the UPS other than
just with computer gear.

And for those of you that like pictures with the explanation:

http://www.explan.co.uk/antisurge/modes.shtml

But apparently it also depends on which country you live and their
wiring standards:

http://www.zerosurge.com/HTML/mode2.html
http://www.zerosurge.com/HTML/works.html
http://www.zerosurge.com/HTML/versus.html

In the U.S., the neutral is bonded to earth ground at the service
entrance, the only way to have common mode surges is from something
inside your house, not from outside. Wouldn't a common mode surge
coming at you from outside become a normal mode surge (since it then
proceeds only on the [hot] line)? I think all the surge arrestors that
I've seen here are mode 1 only (line-to-neutral) because neutral was
tied to local ground (earth) at the service entry.

No, I don't work for ZeroSurge and didn't know about their products
until this thread had me hunting around on surge protection and methods.
Their article at http://www.zerosurge.com/HTML/about.html had me
grinning because I remember calling the telephone company to report
problems. Three times they came out and did their local ring test. No
problems. The fourth time, after looking around myself, I stood there
while they tested. No problems again using both their test gear and
making test calls. Took them to the basement and pointed at the MOV
(metal oxide varistor) used across their lines. Replaced it, problem
fixed. The problems I experienced were indicative of a high resistance
short and I figured the MOV went bad. The MOV has very high resistance
(10^9 ohms) at low voltages and passes very little current, but it has
very low resistance at high voltages which shunts the surge but only for
short-duration impulses. That's why they are rated in Joules (N joules
= N amps across 1 ohm for *1 second*). MOVs can burn out with one good
surge.

Most surge protectors rely on MOVs but these eventually go bad over time
after being repeatedly stressed. Several grains of zinc oxide (each
with an individual breakdown of 3 volts) are sintered together to make a
MOV. Each transient destroys some of those grains which reduces it's
level of protection. The more it has protected you, the less it will
protect you. The eventual failure is a short circuit (which is why
sometimes they'll blow apart) and why a fuse should be used. If not
fused, you'll smell when a MOV eventually shorts; it will smell like an
electrical fire. If MOVs are not themselves fused or enclosed in a fire
retardant, they can start fires (see http://www.smarthome.com/4401.html
for a news report). Get a surge protector that has an LED for status to
tell if the MOV is still good. The MOV should be [thermally] fuse
protected or uses disconnect circuitry to shutoff the surge protector.
That surge protector becomes nothing more than an extension cord when
(not if) the MOV goes bad. A surge arrestor at the mains that goes bad
is worthless, too. I can see how an LED can indicate if a TMOV is still
good (http://snurl.com/2824), presuming the thermal element fries before
the MOV has overly degraded. I suppose for a MOV you could do a similar
setup in using a fuse and tie the LED between them.

I've seen users employ surge protectors and UPS'es to protect the
computer and forget about the telephone line to their modem or the cable
going to their cable modem, especially if an internal modem, or even to
their expensive audio gear. Also, to those that extol installing a
whole-home surge arrestor (which may incur the expense of an electrician
and requires it be your house to toy with), a whole-home surge arrestor
is a point-of-entry device suppressing surges stemming from outside
sources (utility company problems, transformer switching, etc.) but it
won't do anything to suppress the high number of power surges that
originate *inside* your house (appliances, projector bulb burnout,
motors, air conditioning, etc.). I just saw some white papers regarding
the origin of surges but had a problem in IE, had to close a window, and
lost all of them. My recollection was that 20% of surges come from
outside and 80% come from inside. Your A/C causing a surge is not
usually a problem for your computer's power supply, but that huge lathe
in an industrial environment is.

While I can see the arguments for using a point-of-entry or
service-entry surge arrestor for the whole home (if that is really an
option for you, like you own that house), that may not eliminate surges
generated within the home so I still see value for point-of-use surge
protectors. w_tom extols dumping the surge to ground but neglects that
data lines and audio equipment use this as their reference so the surges
that are getting pummeled into local common can cause data and audio
problems. It's a 20-year old spec based on using stand-alone devices
and shunting the surges. Surge voltage on the ground wire varies by its
length. With all the interconnected devices in computer systems, and if
you tie in your audio/video system, too, you will have ground loops. If
the surge induced on the safety ground at one computer is 1000V but it
is 600V at another computer or printer, the 400V difference can cause
disruption, degradation, or damage. Even small surges pumped into the
safety ground wire can cause data errors, lockups, or other "weird"
problems.

 

[Vanguard]

I'm not an electrical engineer but find this topic interesting. Could
you please define what you mean by a "common mode" surge. I thought
that meant the impulse was on both wires (line and neutral) and that a
surge only on 1 wire was a "normal mode" surge. There are probably more
technical explanations of common mode signals but the one at
http://www.explan.co.uk/antisurge/modes.shtml seems to suffice (if you
neglect the resistance to ground which apparently you do in claiming
that slamming the surges to ground does not affect data transmission
which uses ground as the reference).

Say I have a common mode surge headed at my house (same impulse on both
wires). At the service entry, the neutral is grounded to earth. Are
you saying the impulse on the neutral still continues into the house?
Then what was the point of grounding the neutral wire at the service
entrance? And wouldn't this disqualify all your arguments regarding
shunting the surge to ground if the shunt that already and permanently
exists where the neutral wire is bonded to ground at the service entry
doesn't work to eliminate the surge? Why would I expect a surge
arrestor to do any better than a solid heavy gauge wire in shunting the
surge to ground?

If the surge is grounded on the neutral wire and doesn't proceed further
and only the surge on the [hot] line gets through, why would that still
be considered a common mode surge if the [companion] surge is now
missing on the neutral wire?

Also, do you know of any online links to ANSI, IEEE, and UL
specifications (that I don't have to pay for)? As I mentioned, this is
not my career but an interest as a result of my career and hobbies
(i.e., it's not what I do but it can affect what I do). When I have
gone hunting for the specs, it seems I end up as some site that wants me
to pay for the documentation. For example, I want to find out about
UL1449 referenced by a product, I go to http://www.ul.com and click the
link for "Standards", and I get redirected to somewhere else
(http://snurl.com/28fb) where I can buy a copy of the standard for $265.
Yeah, right, like consumers are going to spend more than what the
product costs to find out what the standard means. What the hell good
are any of these standards when they are quoted by manufacturers but the
consumers have to no way to know what they are? And you wonder why
consumers don't know this stuff.

--
____________________________________________________________
** Share with others. Post replies in the newsgroup.
** If present, remove all "-nix" from my email address.
____________________________________________________________

Common mode surge does not become differential once inside
house. It does not because voltage difference remains
between wire with surge and earth ground. That common mode
surge inside the building is still seeking earth ground,
destructively, via appliances.

However common mode surges can induce differential mode
transients. Those long wires bundled with other wires will
induce transients onto those other wires. This is the induced
surge - not from electromagnetic fields of a distant CG
strike. A plug-in protector tries to earth the common mode
surge down safety ground wire. But that long ground wire,
bundled with other wires, only induces the transient on those
other wires. Just another reason why plug-in protectors are
not effective as noted by last paragraph in:
http://www.explan.co.uk/antisurge/modes.shtml

Zerosurge and its competition Brickwall and Surgex are
series mode protectors. Basically just low pass filters.
They cannot work like a dam. Surges will simply overwhelm the
dam. But they do work well as dikes. Difference: a dam must
stop the flow whereas a dike only diverts the flow
downstream. For series mode protectors to work, the surge
must have a downstream path to earth ground. Series mode
protectors will enhance the protection AND should also reduce
noise on that circuit. But, they still require the surge to
be provided a path to earth ground - and that is what shunt
mode protectors do. Shunt mode protectors make a connection
to earth ground only during the surge.

Second reason why series mode protector require that shunt
mode earth ground: the safety ground wire completely bypasses
the filter. If surge is not earthed at service entrance, then
surge will get to electronics by safety ground wire,
unimpeded, right through series mode protector. They forget
to mention that part in the Zerosurge summary. Even series
mode protectors have this requirement - a surge protector is
only as effective as its earth ground.

MOVs do wear out with use. They are useless for small,
daily transients such as spikes from a turned off switch. But
then those spikes are so trivial as to only be called noise.
So small that MOVs don't even see the transients. If those
daily spikes caused damage, then the homeowner was replacing
dimmer switches daily. For audio recording equipment, a
series mode protector would keep those noise spikes away from
recording equipment. But that noise from switches and motors
does not damage electronics. If noise did damage, then we
were all replacing clock radios every day - long before the PC
even existed.

MOV life expectancy is determined by joules. Since most
plug-in protectors do not even claim to provide effective
protection, they are also undersized - too few joules. As
joules increase, the life expectancy of a surge protector
increases exponentially. For example if a 345 joule protector
(only using as little as 115 joules in protection) were to
survive three same size, standard, 8/20 usec surges, then the
1000 joule 'whole house' protector would withstand on the
order of 400 same sized transients. Numbers that plug-in
manufacturers would rather avoid.

Since lightning surges occur typically once every eight
years, then a properly sized 'whole house' protector using
MOVs will last more than long enough. They just forget to
apply numbers when hyping MOV fear in that Zerosurge
discussion.

If using a 'whole house' protector, then any surge, interior
or exterior, is earthed before it can create damage. Nothing
inside the house should create a common mode surge. But even
if something does, the 'whole house' protector still shunts
the common mode transient to earth.

Some papers will outrightly confuse. For example, 80% of
transients can come from internal sources. IOW a 5 volt noise
spike can be generated by an electric motor. If paper is
written by an English major, then that 5 volts suddenly
transpires into a 500 volt surge! In reality, if anything is
creating destructive surges inside a house, then the homeowner
is replacing dimmer switches, clock radios, and GFCI (RCD)
daily. Appliances only create noise - well below what every
household appliance must withstand. Surges come from exterior
events. But too many authors on surges have eyes that glaze
over as soon as numbers are provided. These people see surge
everywhere - which is what plug-in manufacturers live off -
misinformation.

But lets assume that central air conditioner is generating
surges. OK. The surge will immediately take out internal
controls, furnace controls, and the house electronic
thermostat. But if the air conditioner is creating this
surge, then the surge protector must be at the air conditioner
- not on everything else inside the building.

And since central air is a dedicated circuit, then the
'whole house' protector would eliminate that mythical surge.
Air conditioners and other household appliances don't create
surges. They create noise - which hypsters of plug-in
protectors forget to mention since they don't provide (or
learn) numbers.

Want to see an example of outright lies because the author
probably did not understand fundamental concepts and probably
feared numbers?
http://www.howstuffworks.com/surge-protector.htm
Facts so perverted as to be one long chain of lies. Many who
promote surge protector don't even have basic electrical
knowledge such as this author.

Important for the surge protector to be at the service
entrance for exterior common mode surge since those are higher
frequency events. That high frequency in lightning generated
surges are why the connection to earth ground must be less
than 3 meters. At higher frequencies, wire length takes on a
whole different characteristic. It is why a line - neutral
protector is effective when the connection of line to neutral
to earth ground rod is less than 3 meters (and other
electrical characteristics such as not inside metallic
conduit, no sharp bends, etc). Wire length is critical to
where a surge protector is even connected to earth ground.
Earth ground too far away all but does not exist - because of
electrical characteristics of destructive surges.

Your MOV on phone line is why MOVs are no longer used for
such protectors. Semiconductors have long since replaced
those MOVs. They either work, or become short circuits that
tell human to replace them. MOVs have too much capacitance
for use on phone lines. However some pathetic plug-in
protectors do use MOVs for phone line protection. Your
current phone line protector should be semiconductor based as
has been standard for about 15 years. And it should be each
phone wire, through protector, less than 3 meters to single
point ground. Wire to wire protector do comparatively little.

LED to indicate a surge protector as good is really
nonsense. That LED can report one mode of failure, but cannot
report the protector as good. This example from Zerosurge
demonstrates the problem. They even removed all MOVs and the
indicator light still declared the protector as OK:
http://www.zerosurge.com/HTML/movs.html


Transients on internal safety ground wires are irrelevant -
basically don't exist - if the 'whole house' protector is
installed on every incoming utility (or that utility is
directly earthed before entering the building). If using
internal plug-in protectors, then transients would be induced
every where on safety grounds. Described above were induced
surges. But then ground loops would use nearby appliances as
a path to earth ground because breaker box earth ground is too
far away. This is why an adjacent plug-in protector on AC
input to a computer can contribute to damage of a modem - even
if computer is powered off. That is correct. Adjacent
plug-in surge protector completes a circuit to earth ground,
destructively via modern. And modems, especially in North
America, are most often damaged by AC electric surges.

Once a surge has gotten inside the building, then it will
find numerous destructive paths. No plug-in protector is
going to stop that. In fact plug-in protectors can give that
surge more potentially destructive paths to earth ground -
such as the above modem example.

Can MOVs create a fire? That is what UL1449 2nd Edition is
about. Don't know of any ISO equivalent. But then if those
'adjacent to appliance' MOVs were effective, then MOVs are
already inside the appliance. Appliances already have
effective internal protection. Protection that assumes the
'big' transient - the surge - will be earthed before it can
enter the building. Internal protection that makes adjacent
plug-in protectors redundant. Internal protection that makes
any other appliance generated 'noise' (not surge) irrelevant.

Do you need an electrician? Even if plug-in protectors are
installed, the electrician would still be necessary to upgrade
a house ground. Many homeowners can install these 'whole
house' protectors themselves - since it is that easy - and
therefore sold in Home Dept. Even plug-in protectors require
earth grounds. Older homes are too often missing this
connection that is also essential to human safety. In North
America, earth ground should be upgraded to, minimally, post
1990 NEC code requirements. Equivalent earthing is necessary
throughout the world. That AC electric earth ground must be
less than 10 feet from main disconnect (where neutral meets
earth ground) and that earth ground (sometimes called GEC).
And all other incoming utilities must also connect to that
same earth ground.

Exampled was a ground loop problem of computer at 1000 volts
and printer at 400 volts. These problems are created when
single point grounds are not used. Single point grounds that
are necessary and provided by 'whole house' protectors.
Resulting voltage difference is called ground loops. It is
why peripherals connected to a computer should share same
receptacle outlet if not same circuit (or ring). These
transient voltages occur when the surge is not first earthed
at the service entrance. These voltages are not generated by
internal equipment. If any household device generates such
voltages, then either a surge protector must be permanently
connected to that appliance (the absurd solution) or the
device must be removed as a threat to human safety. But then
that device would remove itself; would self destruct anyway.

Devices have more than sufficient internal protection even
on interconnected ports IF the major surges are earthed before
entering the building. Every incoming utility must either
make a direct connection to single point earth ground OR make
that connection via a 'whole house' protector. Plug-in
protectors only provide surges with more destructive paths to
earth ground such as via appliances. Plug-in protectors sell
by forgetting to mention critical earthing OR provide any
useful numbers. Plug-in protectors want you to not even
discuss common mode transients because they are not selling
effective protection. They do not even claim to provide
common mode surge protection.

One fact always remains - a surge protector is only as
effective as its earth ground. No way around that fact.

As regards for *my* personal checklist that says to get a UPS with a
sine wave output, yes, it is unnecessary if all you run from that UPS is
your computer. I know the computer power supply will handle stepped
wave output. Probably the printer, monitor, scanner, speakers, and
modem, too. But eventually you upgrade your system and want a bigger
UPS. In my case, I want a quieter one; the 2kVA unit that I have now
has a fan that makes anywhere from 35 to 50 dB of noise. Will I toss it
because I want one without a fan? No. Will I toss it when the
batteries go bad? No, because they are replaceable; I've already
replaced them once after 5-1/2 years of use (although I decided to use a
bus bar instead of a heavy wire lead) at a cost of $140 for the 2 of
them which was a tenth of buying a new unit. Since I'll use it for
something else, like my audio/video gear, having a sine wave output is
important. It gives me more choices on how to use the UPS other than
just with computer gear.

And for those of you that like pictures with the explanation:

http://www.explan.co.uk/antisurge/modes.shtml

But apparently it also depends on which country you live and their
wiring standards:

http://www.zerosurge.com/HTML/mode2.html
http://www.zerosurge.com/HTML/works.html
http://www.zerosurge.com/HTML/versus.html

In the U.S., the neutral is bonded to earth ground at the service
entrance, the only way to have common mode surges is from something
inside your house, not from outside. Wouldn't a common mode surge
coming at you from outside become a normal mode surge (since it then
proceeds only on the [hot] line)? I think all the surge arrestors that
I've seen here are mode 1 only (line-to-neutral) because neutral was
tied to local ground (earth) at the service entry.

No, I don't work for ZeroSurge and didn't know about their products
until this thread had me hunting around on surge protection and methods.
Their article at http://www.zerosurge.com/HTML/about.html had me
grinning because I remember calling the telephone company to report
problems. Three times they came out and did their local ring test. No
problems. The fourth time, after looking around myself, I stood there
while they tested. No problems again using both their test gear and
making test calls. Took them to the basement and pointed at the MOV
(metal oxide varistor) used across their lines. Replaced it, problem
fixed. The problems I experienced were indicative of a high resistance
short and I figured the MOV went bad. The MOV has very high resistance
(10^9 ohms) at low voltages and passes very little current, but it has
very low resistance at high voltages which shunts the surge but only for
short-duration impulses. That's why they are rated in Joules (N joules
= N amps across 1 ohm for *1 second*). MOVs can burn out with one good
surge.

Most surge protectors rely on MOVs but these eventually go bad over time
after being repeatedly stressed. Several grains of zinc oxide (each
with an individual breakdown of 3 volts) are sintered together to make a
MOV. Each transient destroys some of those grains which reduces it's
level of protection. The more it has protected you, the less it will
protect you. The eventual failure is a short circuit (which is why
sometimes they'll blow apart) and why a fuse should be used. If not
fused, you'll smell when a MOV eventually shorts; it will smell like an
electrical fire. If MOVs are not themselves fused or enclosed in a fire
retardant, they can start fires (see http://www.smarthome.com/4401.html
for a news report). Get a surge protector that has an LED for status to
tell if the MOV is still good. The MOV should be [thermally] fuse
protected or uses disconnect circuitry to shutoff the surge protector.
That surge protector becomes nothing more than an extension cord when
(not if) the MOV goes bad. A surge arrestor at the mains that goes bad
is worthless, too. I can see how an LED can indicate if a TMOV is still
good (http://snurl.com/2824), presuming the thermal element fries before
the MOV has overly degraded. I suppose for a MOV you could do a similar
setup in using a fuse and tie the LED between them.

I've seen users employ surge protectors and UPS'es to protect the
computer and forget about the telephone line to their modem or the cable
going to their cable modem, especially if an internal modem, or even to
their expensive audio gear. Also, to those that extol installing a
whole-home surge arrestor (which may incur the expense of an electrician
and requires it be your house to toy with), a whole-home surge arrestor
is a point-of-entry device suppressing surges stemming from outside
sources (utility company problems, transformer switching, etc.) but it
won't do anything to suppress the high number of power surges that
originate *inside* your house (appliances, projector bulb burnout,
motors, air conditioning, etc.). I just saw some white papers regarding
the origin of surges but had a problem in IE, had to close a window, and
lost all of them. My recollection was that 20% of surges come from
outside and 80% come from inside. Your A/C causing a surge is not
usually a problem for your computer's power supply, but that huge lathe
in an industrial environment is.

While I can see the arguments for using a point-of-entry or
service-entry surge arrestor for the whole home (if that is really an
option for you, like you own that house), that may not eliminate surges
generated within the home so I still see value for point-of-use surge
protectors. w_tom extols dumping the surge to ground but neglects that
data lines and audio equipment use this as their reference so the surges
that are getting pummeled into local common can cause data and audio
problems. It's a 20-year old spec based on using stand-alone devices
and shunting the surges. Surge voltage on the ground wire varies by its
length. With all the interconnected devices in computer systems, and if
you tie in your audio/video system, too, you will have ground loops. If
the surge induced on the safety ground at one computer is 1000V but it
is 600V at another computer or printer, the 400V difference can cause
disruption, degradation, or damage. Even small surges pumped into the
safety ground wire can cause data errors, lockups, or other "weird"
problems.

 

[w_tom]

Welcome to deregulation where we don't want the government
to do standards. As a result, simple standards cost big
bucks.

UL1449 defines testing a surge protector so that it does not
kill humans. Surge protector can fail during some of these
tests and still pass because the only requirement is that
protector not threaten human life. Transistor safety is
irrelevant to UL1449 - or the principles on which all UL
standards are created. Underwriter's Laboratory is about
protecting human life. It does not care whether the product
does as advertised.

Some will cite ANSI/IEEE C62.1 standards. These are
waveforms to simulate surge transients. C62.1 also does not
mean the surge protector does anything. How does a waveform
standard provide surge protection? But again, anything to
sell a plug-in protector. They will hype C62.1 so that the
naive feels this is an effective protector. Too much junk
science reasoning only because a numerical standard is
printed. One should first learn what that standard says.
Most do not when they recommend a surge protector.

A common mode transient inside a building seeks earth
ground. It can seek earth via one or all wires. This being
different from a differential mode transient that requires two
wires. An AC power line transients will be earthed by the
neutral, but other AC wires will continue into the building.
However if the 'whole house' protector earths those other
wires, then common mode surge become irrelevant. However all
incoming wires must have some connection to earth for common
mode transient protection.

Earth has resistance. But more important, wire also has
impedance. Now some numbers. Read multiple times if necessary
since those numbers put surge protection into perspective and
expose junk science myths. For example, a 20 amp copper wire
of 18 meters may appear as 0.2 ohms resistance, but may also
appear as 130 ohms impedance to a surge transient. This wire
*impedance* is why a surge protector must make a less than 3
meter connection to earth.

For example, lets say a 100 amp surge arrives at computer
and adjacent plug-in surge protector. Surge protector
attempts to shunt (divert, connect) that transient to earth
ground via safety ground wire. But that would be 100 amps
times 130 ohms or 13,000 volts. Computer and adjacent surge
protector at something less than 13,000 volts relative to
earth! IOW the surge will find other, destructive paths to
earth such as through modem via phone line. This is typical
modem damage - surge entering on AC mains and leaving to
ground on phone line (UK has a different variation).

Surge arrived on AC hot (black or brown) wire. Adjacent
surge protector shunted surge to green safety ground wire.
Surge protector, being too close to transistor and too far
from earth ground, has contributed to surge damage. Modem
often being a victim because it makes a good conductive path
to earth ground.

Now lets say same 100 amp surge encounters a 'whole house'
protector with a 3 meter earth ground connection. That would
be maybe 4 ohms impedance (and something less than 0.01 ohms
resistance). IOW surge voltage from hot wire to earth ground
is only 400 volts. Voltage at service entrance so low that
1000 and 2000 volt protection abilities built into appliances
are not overwhelmed. Furthermore, that 18 meter wire between
breaker box and appliance NOW becomes part of the surge
protection solution rather than part of the problem.
Impedance in those 18 meters makes it that much more difficult
to find earth ground via the appliance - because surge was
earthed by a 'whole house' protector.

Of course that earth ground rod may not be the best earthing
connection. No matter how big that earth ground, we can never
get an excellent earth ground. Therefore better built homes,
before the foundation is laid, install an Ufer ground with
footings or a halo ground around the entire building. This
makes earth beneath the building equipotential but for cost of
copper wire. Very effective earthing at so little cost.
However that copper wire loop can never be a perfect
conductor; wire has impedance. So we want central earth
ground to be a single point connection and the best earthing
in the facility. Of course we can never make a perfect and
best conductive single point ground.....

Concepts of single point earthing are similar to why Hi-Fi
components require single point grounding to eliminate hum and
why A/D converters in so many electronics have a single point
ground between the analog and digital circuits AT the A/D
converter. Single point ground is a concept that requires
understanding electronic circuit theory and electrical
principles.

These factors combine to make surge damage completely
avoidable. Earthing for no surge damage is so easy if
installed when the building is first built. Even so, we can
install an effective earth ground, in most cases, by superior
single point grounding techniques just at the service
entrance. It may not be fully effective against the 200,000
amp direct strike. But then most people never confront this
rare surge in their lifetime.

My experience demonstrates earthing a direct strike with
only 1 three meter ground rod. Damage to unprotected circuits
was to everything. Damage on the protected circuits -
nothing. Nothing except the utility meter that was partially
damaged and failed later when the surge was earthed through
meter. But this installation was more typical of a best
protected facilities. This one facility has a less than 1
meter connection to a single earth ground rod in conductive
soil. That distance to earth ground determines how much or
little of a surge will seek earth ground inside the building.

As noted previously, once inside the building, a surge will
find too many destructive paths as even demonstrated by an
IEEE paper. Concepts of surge protection were demonstrated in
most papers of the 8 Nov 1998 issue of IEEE Transactions on
Electromagnetic Compatibility. For example, one was a
Norwegian maritime station installed to withstand the EMP of
nuclear war. But so poorly earthed as to be damaged by a
direct lightning strike. Authors demonstrated how earthing
need be repaired.

Legendary application notes are provide by the benchmark in
surge protection - Polyphaser:
http://www.polyphaser.com/ppc_pen_home.asp
Does Polyphaser hype their product line? Of course not.
Polyphaser earns their profits on providing real surge
protection. Most who recommend surge protectors have never
even heard of Polyphaser - which exposes how little they
really knew. Polyphaser discusses earthing - extensively.
Earthing so critical that one product has no earth ground
connection. Instead the protector mounts directly on earth
ground - because distance to earth - 0 meters - is that
critical.

Polyphaser even demonstrates how a direct strike to buried
cable some distance from the building results in transistor
failure - because incoming wire was not properly earthed to a
single point bulkhead:
http://www.polyphaser.com/datasheets/PTD1028.pdf

Lightning strikes somewhere across the street close to the below
grade West cable vault. ... The first line of defense is the
telco protection panel, but the panel must be connected to a low
resistance / inductance ground. There was no adequate ground
available in the telephone room.

Also note that serious surge protectors have little or no
warranty. Polyphaser offers no warranty which is but another
characteristic of better products.

Concepts of earthing for surge protection demonstrated in
this figure from an industry professional. Note that even
buried wires can source an incoming transient. Both structures
have their own central earth ground. All wires connect to
that earth ground before entering structure - either by direct
wire connection or via a surge protector:

http://www.erico.com/erico_public/pdf/fep/TechNotes/Tncr002.pdf

Other concepts:

http://lists.contesting.com/_towertalk/2000-August/032935.html

http://lists.contesting.com/_towertalk/1997-April/004413.html
http://www.psihq.com/iread/ufergrnd.htm
http://scott-inc.com/html/ufer.htm
http://www.psihq.com/iread/strpgrnd.htm
http://www.xantrex.com/support/docserve.asp?id=337
http://www.cinergy.com/surge/ttip08.htm
http://www.ipclp.com/html/aud_ho_faq.html
http://www.telebyteusa.com/primer/ch6.htm (see section 6.4)
http://www.citelprotection.com/citel/tech_article.htm

None of this is mentioned or implied by plug-in protectors
because they are not selling effective surge protection. They
are selling a product that is grossly overpriced (especially
considering how few joules are installed), that avoids all
mention of common mode transients, and that costs on the order
of 10 to 50 times more per protected appliance compared to
'whole house' protectors.

Where do most computer 'experts' get their education on
surge protection? From boxes on retail shelves. Electrical
concepts necessary to understand effective protection are not
taught to nor understood by too many computer experts. For
example, one need not even understand how to use a 3.5 digit
multimeter to get A-plus Certified - to be a certified
computer repairman. Insufficient electrical knowledge is why
so few understand a concept well proven and demonstrated
repeatedly since the 1930s - surge protection.

Surge protectors are not surge protection. Surge protection
is earth ground. A surge protector is only effective when it
makes a connection to earth ground - as those so many
professional comments repeatedly demonstrate. The most
important point that no plug-in protector will ever mention -
a surge protector is only as effective as its earth ground.

I'm not an electrical engineer but find this topic interesting. Could
you please define what you mean by a "common mode" surge. I thought
that meant the impulse was on both wires (line and neutral) and that a
surge only on 1 wire was a "normal mode" surge. There are probably more
technical explanations of common mode signals but the one at
http://www.explan.co.uk/antisurge/modes.shtml seems to suffice (if you
neglect the resistance to ground which apparently you do in claiming
that slamming the surges to ground does not affect data transmission
which uses ground as the reference).

Say I have a common mode surge headed at my house (same impulse on both
wires). At the service entry, the neutral is grounded to earth. Are
you saying the impulse on the neutral still continues into the house?
Then what was the point of grounding the neutral wire at the service
entrance? And wouldn't this disqualify all your arguments regarding
shunting the surge to ground if the shunt that already and permanently
exists where the neutral wire is bonded to ground at the service entry
doesn't work to eliminate the surge? Why would I expect a surge
arrestor to do any better than a solid heavy gauge wire in shunting the
surge to ground?

If the surge is grounded on the neutral wire and doesn't proceed further
and only the surge on the [hot] line gets through, why would that still
be considered a common mode surge if the [companion] surge is now
missing on the neutral wire?

Also, do you know of any online links to ANSI, IEEE, and UL
specifications (that I don't have to pay for)? As I mentioned, this is
not my career but an interest as a result of my career and hobbies
(i.e., it's not what I do but it can affect what I do). When I have
gone hunting for the specs, it seems I end up as some site that wants me
to pay for the documentation. For example, I want to find out about
UL1449 referenced by a product, I go to http://www.ul.com and click the
link for "Standards", and I get redirected to somewhere else
(http://snurl.com/28fb) where I can buy a copy of the standard for $265.
Yeah, right, like consumers are going to spend more than what the
product costs to find out what the standard means. What the hell good
are any of these standards when they are quoted by manufacturers but the
consumers have to no way to know what they are? And you wonder why
consumers don't know this stuff.

 

[Vanguard]

You probably hate being teacher as apparently I'm one of those dumb
students that just doesn't get it. If you want to get off the "hook"
here, any free online resources that I can check that will support your
statements? I see you provided some links and I will get to those
later.

You mentioned that MOVs are no longer used but neglected to mentioned
the "other" solution. A zener diode maybe? Or a pair of high-current
diodes oriented with opposite polarity with a zener in one leg of a
diode to limit the voltage across the lines, like at
http://snurl.com/28kb. Or an overvoltage-triggered SCR? Come on, whet
our appetite and then leave it dangling. I'd sure love to know what's
better than MOVs. I don't like the behavior of MOVs with them degrading
from each use, shorting, and possibly causing a fire hazard. I haven't
the time or budget to purchase a whole bunch of surge protectors,
UPS'es, or power supplies to see what might be used instead. While this
"other" solution would eliminate the degradation problem of MOVs (of
eventually shorting and of catastrophic incendiary sacrifice), the
characteristic fall-over to allow a large current to flow above the
threshold voltage still applies (and still has to be a transient to
prevent too much heat buildup, especially if semiconductors are
involved). Yet you see articles like http://snurl.com/28ke where
varistors are hailed as a better and faster acting solution (for crowbar
or shunt solutions versus series solutions like ZeroSurge proposes that
hopes it can choke the crap out of the transient). Gee, I thought we
were moving away from MOVs.

One should first learn what that standard says.

That's my point. You claim that consumers are dumb. There's a good
reason for this. They cannot get the information! If they have to pay
membership dues and/or the costs to get a copy of the standard then they
might as well as pay some "professional" to do all that for them.
Instead of spending $50 to buy the product, the consumer would have to
pay way over $500 just to get the info to which the product claims it
conforms - and they still haven't bought the product yet! Great idea.
Make standards but hide them from consumers. Consumers are NOT going to
pay to get the information to which a manufacturer professes to conform
so all that gobblety gook printed on the product is WORTHLESS! Yeah, if
it is a product that is sold to consumers that are within the same
industry then it makes sense to publish the standards conformity list on
the product. I doubt that the majority of sales for whole-home surge
arrestors, UPS'es, surge protectors, computers, printers, and the like
are to same-industry insiders. A manufacturer of surge
arrestors/protectors doesn't sell their product to another manufacturer
of surge arrestors/protectors. Nothing on store shelves (retail or
online) is geared for purchase only by 4-year degreed electrical
engineers. If you sell something to someone outside your specialty, you
should provide some decent information to those non-industry consumers.

A common mode transient inside a building seeks earth
ground.

So for a common mode transient to NOT seek ground means it seeks one of
the other lines which means the other line(s) must not also carry the
transient. The reason it seeks ground is that the other lines also
carry the transient so the differential between them isn't any different
than normal; an impulse of 3000V on line (hot) seeks only ground only
because there is also an impulse of 3000V on neutral. That means a
common mode transient has to be on both the line (hot) and neutral so
that the impulse on either will seek ground instead the other line. But
the neutral was grounded at the service entry so it can't have the
impulse. I thought (but don't know where to check) in the U.S. that
surge arrestors operate in mode 1 (line to neutral), not in mode 2 for
L-G and/or N-G.

A common mode transient inside a building seeks earth ground.
It can seek earth via one or all wires. This being
different from a differential mode transient that requires two
wires.

Lost me again. There has to be a potential difference for current to
seek a path across that potential difference. If line (hot) is carrying
a 3000V impulse, and neutral has been bound to ground, then the impulse
on the line has a potential difference with both the neutral and ground
since both are at ground and will seek to both neutral and ground.
Since local common (ground) and neutral are bound to earth ground at the
service entry, the only other "wires" is the line (hot). I'll grant
that neutral and local common are then separated as they are routed
within the building, but by your own admonition the common mode surges
or "noise" generated within a house are of no concern.

An AC power line transients will be earthed by the
neutral, but other AC wires will continue into the building.

Those "wires" then consist of just the line (hot). Since neutral and
local common were bound to earth ground at the service entry, what other
"wires" would carry the transient into the building other than the [hot]
line? I would think that one line (hot) that had the impulse whereas
the other wires (neutral and ground) were earthed do not have the
impulse would be a transverse or normal mode impulse.

Earth has resistance. But more important, wire also has
impedance. ...

Okay, I understand about the impedance. That same impedance back to
earth along the neutral line (since the surge protector should only do
mode 1 crowbarring according to, I think, some 1996-year spec that I
can't read regarding transient suppression) for that 13 kV impulse
shunted at the MOV, or whatever other "solution", would be the same
impedance would be the same impedance for the impulse to travel along
the hot line from the service entry to the point of the crowbar.
However, for sake of argument, we can say the impulse at the service
entrance was 26 kV and got reduced to 13kV at the point of the shunt
from hot to neutral. So the hot and neutral both have the same 13kV
impulse and now have a differential of 0V because of the successful
shunt. So there's no problem for the equipment protected past the
shunt; i.e., the "upstream" equipment got saved.

Surge protector
attempts to shunt (divert, connect) that transient to earth
ground via safety ground wire.

No, mode 1 surge protection shunts to neutral, not [safety] ground.
Surge protectors that do all-mode (L-N, L-G, N-G) are dangerous since
shunting to safety ground defeats the *safety* of the ground. Someone
touching the metal case which is attached to ground when the impulse got
shunted from hot to ground and who also happened to be touching a pipe
or other heavy conduit to earth would "feel the surge." And since there
is a sizable albeit short-lived current. It's not like a static shock
that has almost no current and where you simply jerk from the shock.
Shunting to safety ground poses a health hazard. You don't put
"protectors" in a home that can kill people, which, as you say, is the
purpose of the UL standards to protect life and limb, and property
survival is secondary. I think that was also part of the 1996-year
transient standard (that I cannot check on) where only mode 1 is
allowed.

You do, however, now have 2 lines (hot and neutral) that have a 13 kV
potential difference to the safety ground as opposed to the 1 line (hot)
that you had before.

IOW
?

the surge will find other, destructive paths to
earth such as through modem via phone line.

Yeah, for the equipment that is not on the output side of the surge
protector (i.e., upstream of the shunt). The shunt protects upstream
equipment, not downstream equipment. You can buy surge protectors with
more than enough outlets in which to plug all your computer equipment;
if not, you can add powerstrips upstream of the surge protector. If you
use 2 different surge protectors, one for your computer gear and another
for your audio/video equipment, and despite them being plugged into the
same wall outlet, the effective impulse at the shunt in each surge
protector can be different by several hundred volts (due to the
impedance from the length of the wires from the shunt in one surge
protector to the shunt in the other surge protector). So obviously you
don't want to have a connection running from your sound card to your
stereo or from your video card's TV out to your television when these
systems are on different surge protectors. Those interconnections are
okay if you get a surge protector with more outlets so all of that
equipment is on the same surge protector, or use a powerstrip upstream
of the surge protector to put all your computer and audio/video gear on
the same surge protector so they are all past the shunt for that surge.

If you have modems getting fried, my bet is that the large A/C-D/C
converter on the end of the power cord for the modem (which is the power
supply) was too big for the surge protector's outlets or covered up too
many other outlets so the user plugged it into the wall or into another
power strip (surge protected or not) that was plugged into the wall.
One side of the modem was connected to the protected computer system
which doesn't experience the surge. The other side of the modem does
experience the surge. You protected one side of the modem but not the
other. Same goes for the telephone line input to the modem. You need
to use a surge protector that incorporates a port for the telephone line
so the surge from the telephone line also gets shunted before it reaches
the modem AND without getting fed back to the computer (to then surge
the RS-232 or PCI slot side of the modem). Surge protectors that simply
shunt across the telephone line separately of the shunting for the power
lines are no different than using 2 different surge protectors to
protect 2 groups of separated equipment and then interconnecting that
equipment across the 2 surge protectors as in the computer to A/V
scenario.

You don't go stringing surge capacitors from every wall outlet to
protect that upstream equipment and then interconnect the upstream
equipment across the surge protectors. Well, I suppose consumers
probably do because the surge protectors probably didn't tell them not
to, but then they also spewed a bunch of standards on the printed label
on their product without the consumers having a means of deciphering
what all that gibberish means.

'whole house' protector

I wholeheartedly agree with you (or maybe it's "I vehemently agree with
you" in this case) that a whole-home surge arrestor is the best
solution - if it is an option to you. If you own the property, you
probably are not restricted from altering that property. If you rent
the house, you can't change the wiring even to add safety equipment
without permission. If you are in an apartment, you usually have access
only to the circuit breaker panel and that's it, and you are not
permitted to even change anything there. If you can do it (to make the
change) and if can do it (to install the whole-home surge arrestor), I'd
say go for it right now. If you can't do the install yourself but are
willing to pay an electrician then plan your budget to get that done as
soon as possible. But if you canNOT do it (because it's not your
property) then you have to find other solutions. One is to use surge
protectors - but never interconnect the upstream equipment attached to
separate surge protectors, and never connect some of the gear to the
surge protector but leave part of it connected downstream of the surge
protector, like the modem's honking big A/C-D/C converter or its
telephone line. If you use multiple surge protectors, make sure it
floats on the upstream side of that surge protector without ever
connecting to anything attached on a different surge protector.

You argue what is best. I'm arguing what is doable. If best is doable,
do that. If best is not doable, do something lesser but still do it.

Concepts of single point earthing are similar to why Hi-Fi
components require single point grounding to eliminate hum and
why A/D converters in so many electronics have a single point
ground between the analog and digital circuits AT the A/D
converter.

Or why you end up using an isolation transformer to break the ground
loop between preamp or low-level inputs and amplifier. Or why you use a
1-to-1 isolation transformer around a ferite core to isolate an antenna
input on a receiver with its ground from the distanced ground for the
antenna amplifier.

Single point ground is a concept that requires
understanding electronic circuit theory and electrical
principles.

I don't doubt it. But the non-industry public never really gets much
education on educational television. And apparently trying to get at
the standards set by those industries is overly improbable to the
consumers that use those products produced by those industries. The
industry isn't very forthcoming with detailed information regarding
their products, either. Ever try to get the actual diagrams for a surge
protector, UPS, or even a power supply? Even if they don't have any
patented circuits or gizmos, they still hide what *is* their product.
And if it's patented then it's patented and they should still be able to
tell you. A lot of fear. But then a lot of thiefs, too.

Thanks for all the links. I'll go visit them and see what else I can
learn or, at least, get an opinion to see if it can be verified by
another.
(The http://www.polyphaser.com/datasheets/PTD1028.pdf link didn't work
so I'll have to hunt around on that site for it.)

Also note that serious surge protectors have little or no
warranty.

Well, the warranties [extended to connected equipment] are often there
but the real burden is proving that the damage was caused by a surge
that got past their equipment, and then usually having to re-prove it to
a judge. With a vendor of a graphic card, the cost of supplying a
replacement is very minor. But having to settle on a $25K to $50K claim
on a warranty means you end up in court, or if you settle out of court
then, as with all such settlements, you waive your right to divulge any
information regarding the settlement (i.e., if they pay, you have to
shut up about the whole mess).

Polyphaser offers no warranty which is but another
characteristic of better products.

Doesn't matter, however, since Uniform Commercial Codes and state
warranty laws will enforce intent of purpose. You sell a surge
protector then that's what you are selling and it better had do its job,
but the only coverage for loss in that case is only for the product and
nothing connected to it. Warranties only provide coverage in addition
to what the law requires, and whether stated or not they will recognize
that "laws in your area may apply which affect coverage and enforcement
of this warranty and/or other implied warranties" or something along
that line. Just because you don't provide a warranty doesn't mean there
isn't one. That's the first time someone has claimed that the *lack* of
a warranty equates to *greater* quality.

"We sell it. We don't stand behind anything we sell."
"Everything is ... AS-IS."
"We make promises about what it will do. We're good at making promises.
We don't promise to honor those promises."

I think what you meant to say is that Polyphaser only warranties its own
products and nothing that is connected to it or anything generated from
that connected equipment. They don't provide a "lure" to bait
consumers, a lure that almost instantly disintegrates when you try to
exercise that warranty. But that's no big surprise. Microsoft Excel's
warranty only warranties the software and the only recourse you have is
to get your money back for the cost of its purchase, not for all the
data it screwed up. Very few warranties ever cover anything other than
the product itself without regard to whatever else the failure of the
product might have affected.

Thanks for all the info. More meat to chew on.

 

[w_tom]

MOVs are not used on phone lines - too much capacitance.
Semiconductor based devices are used; replacing a MOV like
device often called the carbons:
http://www.inwap.com/inwap/chez/Phoneline.jpg
Another is called Sidactor. Avalanche diodes (Transzorb,
Transil) are sometimes used as low power replacements for
MOVs. But conventional avalanche diodes also have too much
capacitance for phones. Therefore specialized version is used.

For conventional power protection, there is no better 'buy'
than MOVs. If worried about degradation or fire, then MOVs
are insufficient and undersized - as found in some power
strips. Fire is no longer a problem if the surge protector is
UL1449 certified AND located where fire is not a danger such
as the service entrance. But where does a plug-in surge
protector end up? Under papers on a desk or on floor behind
furniture in a deep shag carpet or in a pile of dust balls.
UL1449 says they should not create sparks or fire, but is that
where you want an undersized protector?

As stated previously, series mode surge protectors such as
Zerosurge, et al still require the earthing provided by
shunt mode protectors. Complete facts not provided in that
Zerosurge discussion of MOVs forget to mention that MOVs are
some of the best surge protectors. That fire and degradation
problem only exists when a myth purveyor forgets to provide
numbers - promotes using junk science. Plug-in protector does
have those problems when undersized.

What good is providing facts? Toothpaste did nothing until
Crest provided Fluoride. Listerene still does nothing. And
yet so many will promote and buy Listerene and Scope as
demonstrated by the large shelf space in supermarkets. Why
don't they provide the numbers? Because profits are too great
to be fully honest about mouthwash - and how it does nothing
effective. Consumers often buy on feel and on advertisement
that promotes half facts. Or STP oil treatment. Or Oxyclean
detergent. Needed are more Consumer Reports.

Problem is that once a person had accepted something as fact
(believes the propaganda without much doubting), then he holds
all contrarian information to a higher threshold of
acceptance. He outrightly doubts anything contrary to that
original propaganda. For example, the public thinks the
blackout was a result of insufficient capacity. Nonsense. The
grid was not fully loaded - had sufficient capacity - when the
blackout happened. NERC and DOE state that capacity is not a
problem - now or in the projected future. Problem is
control. The grid does not have a facility shortage. It has
a serious control problem directly traceable to how some
company management (ie FirstEnergy) operate. But since early
pundits said the grid is too old, too overloaded, and not
built to do what it is now doing, then the public only hears
and believes that propaganda. Propaganda - not the facts
published before the blackout by NERC and DOE - are what the
public hears.

This same mindset is also why plug-in surge protectors are
sold. People are quick to only believe what they first heard
- even though scientific facts - the truth - takes longer to
be
discovered, presented, and explained.

However information printed on a surge protector is not
gobblety gook. Missing facts demonstrate that plug-in
protectors are not effective. They will claim differential
mode protection. But they don't claim common mode
protection. They don't even discuss critical earthing.
However a public that believes a surge protector will stop,
block, or absorb a surge cannot be bothered with the new
discovery - missing information.

Computer power supplies are also sold using the same
deception. All power supplies can be shorted and not
damaged. Supply can fail (regulator outputs excessive
voltage) and nothing inside a computer must experience that
excessive voltage (overvoltage protection). If the computer
attempts to draw too much power, then a power supply is not
damaged (overpower protection). However so many buy supplies
without ever looking for specifications. Therefore $20 power
supply (that also create radio interference) are sold widely -
because the computer assembler never learned basic electricity
and makes no attempt to do so. The information is in
specifications - most often in the missing parts of that
specification. That $20 supply never provides specifications
because it does not provide essential power supply functions.

So many ineffective plug-in surge protectors don't even
provide those specifications - and that, unfortunately, is
acceptable to any consumers.

The junk scientist says his computer works today, therefore
a power supply is not defective. But a power supply missing
those essential functions really is defective - regardless of
whether the computer works today.

How all technologies work. Welcome to a problem in America
- too many business degrees and too little education on how
things work. Too many feel that don't need to see the bigger
picture because computer works today. Myopia of business
school graduates as even W E Deming repeatedly noted.

Fundamental to understanding how things work: it must work
both experimentally and theoretically - else it does not
work. That concept is taught in high school science. And yet
still people will deny their education - using junk science
reasoning instead, such as a big buck warranty, to believe
facts not in existence. Because the power supply works now,
then it is not defective? Because Shuttle launched in
freezing weather previously, then it will always launch in
freezing weather without exploding? That was the reasoning.
Junk science reasoning is widespread.


Differential mode transient - two wires where transient only
seeks that second wire as a return path - nothing else.
Common mode - one or many wires where the transient can seek
any path to earth ground. All surge protectors install MOVs
L-N, L-G, and N-G - what was described in last post as
dangerous.

Shunting to safety ground poses a health hazard. You don't
put "protectors" in a home that can kill people...

But that is how plug-in protector are wired. And that short
transient is not considered dangerous to people according to
UL standards. Person must first be part of a complete
circuit. Incoming on grounded appliance, but where is the
outgoing path? Destructive to appliance but not necessarily
dangerous to human.

Again, wire impedance means that plug-in protector will not
shunt a common mode surge only back to breaker box. It
will shunt that surge also destructively in other directions
through appliance - and yes, maybe through a person. (But
only if that person has a different good connection to earth
ground.)

Also if the surge was differential mode, then the surge
would never seek those other, destructive paths.

Common mode surge: cut neutral and ground wire; and the
common mode surge will still seek earth ground only down the
remaining hot wire. Surge continues unaffected. Cut that
neutral and ground wire; the differential mode surge current
cannot exist.

I don't understand your 13K (and 26K which does not exist)
example. But assumed that transient is a voltage source.
Current source is why impedance on black wire (coming to
appliance) is irrelevant AND why series mode protectors are
not effective without corresponding shunt mode protection.
Volt will increase, as necessary, to overwhelm any impedance
on the black wire AND created by low pass filter in a series
mode protector. Because the transients is current mode, then
a transient must be earthed before it can enter the building.
Else voltage will increase, as necessary, to overwhelm series
mode protector or galvanic (isolation transformer) protector,
or internal appliance protection.

Furthermore a plug-in protector does try to shunt the surge
on safety ground. Yes it may also shunt on neutral, but
plug-in protector are not effective because they do shunt
surges on safety ground - a distance too far from earth
ground.


Modem damage. Does not matter whether modem was internal or
external. Common mode transient from AC mains found earth
ground, destructively, via modem.

Lets assume plug-in surge protector is connected to same
outlet that an another electronic appliance takes power. Some
think the appliance is not protected if not plugged into the
power strip. Wrong. The appliance in outlet gets same shunt
protection as if plugged into power strip. If power strip
surge protector stops or blocks surges, then an appliance
would need be plugged into the power strip. But that is not
what shunt mode surge protectors (ie power strip protectors)
do. They shunt meaning that everything around them - before
and after the power strip - have equivalent protection.


Lets say you don't own the property - cannot get landlord to
install a 'whole house' protector. Best one can do is locate
an outlet closest to breaker box (on same phase) and plug the
largest joule plug-in protector into that outlet. Best place
for that plug-in protector may be outlet on breaker box - the
closest connection to earth ground. To install a plug-in
protector as if it was a 'whole house' protector. This
assumes the building has some earth ground. Many, even built
in 1970, are missing that essential (and once provided) safety
ground. No earth ground means no surge protector will be
effective. Pragmatism - good luck getting a landlord to
upgrade his earthing system since so many say, "Lights work
just fine" and will not spend money. As demonstrated by the
previous power supply and Shuttle examples, the landlord is
denying failure exists.

To complicate issues, those three light outlet testers will
report a defective ground. But that does not test earth
ground. It is only a safety ground check. Some actually
have posted that an outlet checker will verify earth ground
AND that earth ground is tested by a ground indicator light on
some power strips. Those tests will detect a failed safety
ground, but will not report a good safety ground, and make no
claims for earth ground.

Actually, the best solution to surge protection would be to
have it required on all homes - as is required on phone
lines. First a halo ground or Ufer ground would be installed
when home is built. Or the concrete basement floor using the
reinforcement wire mesh as part of the earthing system - to
make earth beneath a building equipotential. The 'whole
house' protector would be installed by same electrical code
rules that solved electric fire problems. Then we have an
earthing system that even solves surge problems created by
geological variations AND we have a surge protector that is
already properly connected.

Currently we must customize every surge protector
installation because we don't install earth ground and surge
protectors in a standard installation. Customizing makes the
'how' part complex. This need for surge protectors is quite
recent - only with the 1970 appearance of transistors in all
homes. However we still build new homes as if the transistor
did not exist. A standard installation including superior
earthing and surge protector would provide almost universal
protection at very low cost. Presently, some homes need
massive earthing (that was not as expensive when the house was
being built) to solve future surge damage. All surge damage
is that avoidable and unnecessary.

You probably hate being teacher as apparently I'm one of those dumb
students that just doesn't get it. If you want to get off the "hook"
here, any free online resources that I can check that will support your
statements? I see you provided some links and I will get to those
later.

You mentioned that MOVs are no longer used but neglected to mentioned
the "other" solution. A zener diode maybe? Or a pair of high-current
diodes oriented with opposite polarity with a zener in one leg of a
diode to limit the voltage across the lines, like at
http://snurl.com/28kb. Or an overvoltage-triggered SCR? Come on, whet
our appetite and then leave it dangling. I'd sure love to know what's
better than MOVs. I don't like the behavior of MOVs with them degrading
from each use, shorting, and possibly causing a fire hazard. I haven't
the time or budget to purchase a whole bunch of surge protectors,
UPS'es, or power supplies to see what might be used instead. While this
"other" solution would eliminate the degradation problem of MOVs (of
eventually shorting and of catastrophic incendiary sacrifice), the
characteristic fall-over to allow a large current to flow above the
threshold voltage still applies (and still has to be a transient to
prevent too much heat buildup, especially if semiconductors are
involved). Yet you see articles like http://snurl.com/28ke where
varistors are hailed as a better and faster acting solution (for crowbar
or shunt solutions versus series solutions like ZeroSurge proposes that
hopes it can choke the crap out of the transient). Gee, I thought we
were moving away from MOVs.

One should first learn what that standard says.

That's my point. You claim that consumers are dumb. There's a good
reason for this. They cannot get the information! If they have to pay
membership dues and/or the costs to get a copy of the standard then they
might as well as pay some "professional" to do all that for them.
Instead of spending $50 to buy the product, the consumer would have to
pay way over $500 just to get the info to which the product claims it
conforms - and they still haven't bought the product yet! Great idea.
Make standards but hide them from consumers. Consumers are NOT going to
pay to get the information to which a manufacturer professes to conform
so all that gobblety gook printed on the product is WORTHLESS! Yeah, if
it is a product that is sold to consumers that are within the same
industry then it makes sense to publish the standards conformity list on
the product. I doubt that the majority of sales for whole-home surge
arrestors, UPS'es, surge protectors, computers, printers, and the like
are to same-industry insiders. A manufacturer of surge
arrestors/protectors doesn't sell their product to another manufacturer
of surge arrestors/protectors. Nothing on store shelves (retail or
online) is geared for purchase only by 4-year degreed electrical
engineers. If you sell something to someone outside your specialty, you
should provide some decent information to those non-industry consumers.

A common mode transient inside a building seeks earth
ground.

So for a common mode transient to NOT seek ground means it seeks one of
the other lines which means the other line(s) must not also carry the
transient. The reason it seeks ground is that the other lines also
carry the transient so the differential between them isn't any different
than normal; an impulse of 3000V on line (hot) seeks only ground only
because there is also an impulse of 3000V on neutral. That means a
common mode transient has to be on both the line (hot) and neutral so
that the impulse on either will seek ground instead the other line. But
the neutral was grounded at the service entry so it can't have the
impulse. I thought (but don't know where to check) in the U.S. that
surge arrestors operate in mode 1 (line to neutral), not in mode 2 for
L-G and/or N-G.

A common mode transient inside a building seeks earth ground.
It can seek earth via one or all wires. This being
different from a differential mode transient that requires two
wires.

Lost me again. There has to be a potential difference for current to
seek a path across that potential difference. If line (hot) is carrying
a 3000V impulse, and neutral has been bound to ground, then the impulse
on the line has a potential difference with both the neutral and ground
since both are at ground and will seek to both neutral and ground.
Since local common (ground) and neutral are bound to earth ground at the
service entry, the only other "wires" is the line (hot). I'll grant
that neutral and local common are then separated as they are routed
within the building, but by your own admonition the common mode surges
or "noise" generated within a house are of no concern.

An AC power line transients will be earthed by the
neutral, but other AC wires will continue into the building.

Those "wires" then consist of just the line (hot). Since neutral and
local common were bound to earth ground at the service entry, what other
"wires" would carry the transient into the building other than the [hot]
line? I would think that one line (hot) that had the impulse whereas
the other wires (neutral and ground) were earthed do not have the
impulse would be a transverse or normal mode impulse.

Earth has resistance. But more important, wire also has
impedance. ...

Okay, I understand about the impedance. That same impedance back to
earth along the neutral line (since the surge protector should only do
mode 1 crowbarring according to, I think, some 1996-year spec that I
can't read regarding transient suppression) for that 13 kV impulse
shunted at the MOV, or whatever other "solution", would be the same
impedance would be the same impedance for the impulse to travel along
the hot line from the service entry to the point of the crowbar.
However, for sake of argument, we can say the impulse at the service
entrance was 26 kV and got reduced to 13kV at the point of the shunt
from hot to neutral. So the hot and neutral both have the same 13kV
impulse and now have a differential of 0V because of the successful
shunt. So there's no problem for the equipment protected past the
shunt; i.e., the "upstream" equipment got saved.

Surge protector
attempts to shunt (divert, connect) that transient to earth
ground via safety ground wire.

No, mode 1 surge protection shunts to neutral, not [safety] ground.
Surge protectors that do all-mode (L-N, L-G, N-G) are dangerous since
shunting to safety ground defeats the *safety* of the ground. Someone
touching the metal case which is attached to ground when the impulse got
shunted from hot to ground and who also happened to be touching a pipe
or other heavy conduit to earth would "feel the surge." And since there
is a sizable albeit short-lived current. It's not like a static shock
that has almost no current and where you simply jerk from the shock.
Shunting to safety ground poses a health hazard. You don't put
"protectors" in a home that can kill people, which, as you say, is the
purpose of the UL standards to protect life and limb, and property
survival is secondary. I think that was also part of the 1996-year
transient standard (that I cannot check on) where only mode 1 is
allowed.

You do, however, now have 2 lines (hot and neutral) that have a 13 kV
potential difference to the safety ground as opposed to the 1 line (hot)
that you had before.

IOW
?

the surge will find other, destructive paths to
earth such as through modem via phone line.

Yeah, for the equipment that is not on the output side of the surge
protector (i.e., upstream of the shunt). The shunt protects upstream
equipment, not downstream equipment. You can buy surge protectors with
more than enough outlets in which to plug all your computer equipment;
if not, you can add powerstrips upstream of the surge protector. If you
use 2 different surge protectors, one for your computer gear and another
for your audio/video equipment, and despite them being plugged into the
same wall outlet, the effective impulse at the shunt in each surge
protector can be different by several hundred volts (due to the
impedance from the length of the wires from the shunt in one surge
protector to the shunt in the other surge protector). So obviously you
don't want to have a connection running from your sound card to your
stereo or from your video card's TV out to your television when these
systems are on different surge protectors. Those interconnections are
okay if you get a surge protector with more outlets so all of that
equipment is on the same surge protector, or use a powerstrip upstream
of the surge protector to put all your computer and audio/video gear on
the same surge protector so they are all past the shunt for that surge.

If you have modems getting fried, my bet is that the large A/C-D/C
converter on the end of the power cord for the modem (which is the power
supply) was too big for the surge protector's outlets or covered up too
many other outlets so the user plugged it into the wall or into another
power strip (surge protected or not) that was plugged into the wall.
One side of the modem was connected to the protected computer system
which doesn't experience the surge. The other side of the modem does
experience the surge. You protected one side of the modem but not the
other. Same goes for the telephone line input to the modem. You need
to use a surge protector that incorporates a port for the telephone line
so the surge from the telephone line also gets shunted before it reaches
the modem AND without getting fed back to the computer (to then surge
the RS-232 or PCI slot side of the modem). Surge protectors that simply
shunt across the telephone line separately of the shunting for the power
lines are no different than using 2 different surge protectors to
protect 2 groups of separated equipment and then interconnecting that
equipment across the 2 surge protectors as in the computer to A/V
scenario.

You don't go stringing surge capacitors from every wall outlet to
protect that upstream equipment and then interconnect the upstream
equipment across the surge protectors. Well, I suppose consumers
probably do because the surge protectors probably didn't tell them not
to, but then they also spewed a bunch of standards on the printed label
on their product without the consumers having a means of deciphering
what all that gibberish means.

'whole house' protector

I wholeheartedly agree with you (or maybe it's "I vehemently agree with
you" in this case) that a whole-home surge arrestor is the best
solution - if it is an option to you. If you own the property, you
probably are not restricted from altering that property. If you rent
the house, you can't change the wiring even to add safety equipment
without permission. If you are in an apartment, you usually have access
only to the circuit breaker panel and that's it, and you are not
permitted to even change anything there. If you can do it (to make the
change) and if can do it (to install the whole-home surge arrestor), I'd
say go for it right now. If you can't do the install yourself but are
willing to pay an electrician then plan your budget to get that done as
soon as possible. But if you canNOT do it (because it's not your
property) then you have to find other solutions. One is to use surge
protectors - but never interconnect the upstream equipment attached to
separate surge protectors, and never connect some of the gear to the
surge protector but leave part of it connected downstream of the surge
protector, like the modem's honking big A/C-D/C converter or its
telephone line. If you use multiple surge protectors, make sure it
floats on the upstream side of that surge protector without ever
connecting to anything attached on a different surge protector.

You argue what is best. I'm arguing what is doable. If best is doable,
do that. If best is not doable, do something lesser but still do it.

Concepts of single point earthing are similar to why Hi-Fi
components require single point grounding to eliminate hum and
why A/D converters in so many electronics have a single point
ground between the analog and digital circuits AT the A/D
converter.

Or why you end up using an isolation transformer to break the ground
loop between preamp or low-level inputs and amplifier. Or why you use a
1-to-1 isolation transformer around a ferite core to isolate an antenna
input on a receiver with its ground from the distanced ground for the
antenna amplifier.

Single point ground is a concept that requires
understanding electronic circuit theory and electrical
principles.

I don't doubt it. But the non-industry public never really gets much
education on educational television. And apparently trying to get at
the standards set by those industries is overly improbable to the
consumers that use those products produced by those industries. The
industry isn't very forthcoming with detailed information regarding
their products, either. Ever try to get the actual diagrams for a surge
protector, UPS, or even a power supply? Even if they don't have any
patented circuits or gizmos, they still hide what *is* their product.
And if it's patented then it's patented and they should still be able to
tell you. A lot of fear. But then a lot of thiefs, too.

Thanks for all the links. I'll go visit them and see what else I can
learn or, at least, get an opinion to see if it can be verified by
another.
(The http://www.polyphaser.com/datasheets/PTD1028.pdf link didn't work
so I'll have to hunt around on that site for it.)

Also note that serious surge protectors have little or no
warranty.

Well, the warranties [extended to connected equipment] are often there
but the real burden is proving that the damage was caused by a surge
that got past their equipment, and then usually having to re-prove it to
a judge. With a vendor of a graphic card, the cost of supplying a
replacement is very minor. But having to settle on a $25K to $50K claim
on a warranty means you end up in court, or if you settle out of court
then, as with all such settlements, you waive your right to divulge any
information regarding the settlement (i.e., if they pay, you have to
shut up about the whole mess).

Polyphaser offers no warranty which is but another
characteristic of better products.

Doesn't matter, however, since Uniform Commercial Codes and state
warranty laws will enforce intent of purpose. You sell a surge
protector then that's what you are selling and it better had do its job,
but the only coverage for loss in that case is only for the product and
nothing connected to it. Warranties only provide coverage in addition
to what the law requires, and whether stated or not they will recognize
that "laws in your area may apply which affect coverage and enforcement
of this warranty and/or other implied warranties" or something along
that line. Just because you don't provide a warranty doesn't mean there
isn't one. That's the first time someone has claimed that the *lack* of
a warranty equates to *greater* quality.

"We sell it. We don't stand behind anything we sell."
"Everything is ... AS-IS."
"We make promises about what it will do. We're good at making promises.
We don't promise to honor those promises."

I think what you meant to say is that Polyphaser only warranties its own
products and nothing that is connected to it or anything generated from
that connected equipment. They don't provide a "lure" to bait
consumers, a lure that almost instantly disintegrates when you try to
exercise that warranty. But that's no big surprise. Microsoft Excel's
warranty only warranties the software and the only recourse you have is
to get your money back for the cost of its purchase, not for all the
data it screwed up. Very few warranties ever cover anything other than
the product itself without regard to whatever else the failure of the
product might have affected.

Thanks for all the info. More meat to chew on.

 

[w_tom]

Looking back on your original list of damaged components, I
notice PSU was damaged. What type of PSU was this? What /
where are specifications for this PSU?

 

[David LeBrun]

Not sure...but I'm told it was a cheap one...and was replaced with an
Enermax 300W
I don't know if the PSU was replaced before the rest of the damage
occured or if it all happened at the same time.

I have been following along with the discussion but most of what you
and Vanguard have been exchanging is beyond me.

Dave.

 

[w_tom]

Here is the reason I asked about supply. Was there a power
surge or was a power surge just assumed? Original power
supply was cheap - often missing essential overvoltage
protector circuits. If this inferior power supply suffered
regulator failure, then power supply outputs excessive and
destructive voltages. In a standard power supply such as the
Enermax, overvoltage protector circuits would have shorted
out that overvoltage. Other components inside the system
would not have been damaged because overvoltage was shorted
out by protector circuit - that is often missing in cheap
power supplies.

Maybe there was no power surge. Damage was simply created
by a human who did not know to get a supply with essential
functions such as that overvoltage protector. IOW supply was
purchased only on price; not using specifications. Therefore
the supply failed and took out that long list of computer
components.

Many of the damaged components are typical of a power supply
sold with a good price and no specifications. Only component
not damaged that I would have expected damage to was RAM.
Power supply would destroy disk drives but not CPU.

IOW that above entire discussion may have been irrelevant
because system was not damaged by a surge. System was damaged
by human who did not know how to buy a power supply. Such
defective supplies are widespread; even sold in CompUSA.

 

[William]

Howdy there, a good thread. I have used Enermax power supplies in the
past and have been very pleased with them. I also have an Antec power
supply that came with an Antec case I recieved. Both power supply units
have been excellent. From my experience, you get what you pay for. If
you want a good quality component, I suggest you purchase products
accordingly ! There are other excellent manufacturers of power supply
units out there as well. Just be careful what you buy. I suggest you
leave it to those who have had many year experience to give advice on
which manufacturers have good reputations and courteous service with their
products. Shall the best come out of your endeavor.

 

[David LeBrun]

When the system came to me I was "told" it was a power surge that
destroyed the components. I really don't know if this is the case and
given all that was said in this thread I would likely discount this as
the cause. I visually inspected all the damaged components but I
didn't see any signs except for the sound card with the popped core.
I was looking for bulging/leaking caps, discolored/chipped/missing
parts but could not find anything. I ran some burn-in tests on the
remaining parts and nothing came up as bad.

I do remember being told by the owner that one person doing work on
the unit flipped a switch on the PSU...however she doesn't remember it
having the on/off switch that the current Enermax does...so I can only
assume the switch would be the voltage selector. Now...I have never
operated a unit when it was not set for North American voltage (110 or
whatever its supposed to be). Another possibility I can think of is
that someone did work on the board when it still had power...full or
standby...but even then I don't think that much damage would occur.
I've oops'd like that a couple of times and the board simply shut
down...although that was with an Enermax so who knows what an el'
cheapo unit would do. I just find it really strange which components
died and those that survived. Another thing I just realized is that
nothing else in her house was damaged...not even the monitor which was
on the same outlet/power bar...so it most likely was the PSU and/or
human error. Anyways...the owner purchased new parts and has her
system back (minus her data) with a better power supply.

 

[kony]

When the system came to me I was "told" it was a power surge that
destroyed the components. I really don't know if this is the case and
given all that was said in this thread I would likely discount this as
the cause. I visually inspected all the damaged components but I
didn't see any signs except for the sound card with the popped core.
I was looking for bulging/leaking caps, discolored/chipped/missing
parts but could not find anything. I ran some burn-in tests on the
remaining parts and nothing came up as bad.

I do remember being told by the owner that one person doing work on
the unit flipped a switch on the PSU...however she doesn't remember it
having the on/off switch that the current Enermax does...so I can only
assume the switch would be the voltage selector. Now...I have never
operated a unit when it was not set for North American voltage (110 or
whatever its supposed to be). Another possibility I can think of is
that someone did work on the board when it still had power...full or
standby...but even then I don't think that much damage would occur.
I've oops'd like that a couple of times and the board simply shut
down...although that was with an Enermax so who knows what an el'
cheapo unit would do. I just find it really strange which components
died and those that survived. Another thing I just realized is that
nothing else in her house was damaged...not even the monitor which was
on the same outlet/power bar...so it most likely was the PSU and/or
human error. Anyways...the owner purchased new parts and has her
system back (minus her data) with a better power supply.

It probably was a power surge, the core on a sound card doesn't
generally do that from any kind of power supply failure or incorrect
voltage switch, etc, etc.


Dave

 

[w_tom]

It is a sign of the times; how propaganda manipulates
mindsets. Today, any failure must have been a surge.
Computer mysteriously turns off; must be a surge. However
surges occur typically once every eight years. Propaganda
forgot to mention that part.

That list of failed components is consistent with power
supply failure that did not include the required overvoltage
protection circuits. Circuits that were even standard 30
years ago are commonly missing on power supplies sold only on
price. However specifications for an Enermax power supply
states overvoltage protection - with numbers:
http://www.enermax.com.tw/product.htm
Fundamental difference between the power supply purchased on
price verses power supply bought on value.

Many discount power supplies cannot be bothered to cite
specifications because they are not selling to those who want
essential power supply functions. List of damaged components
is consistent with a power supply that did not have
overvoltage protection - was sold only on one spec - price.

 

[Vanguard]

I finally got some response to e-mails that I sent to my cable broadband
provider and my telephone company. Neither one provides surge
protection (at the service entry or anywhere else). UFFDA! So much for
the telco providing surge protection on their service, or having any
protection on the cable coax. I'm still reading through all the links
you provided (had to do some house restoration since then and been
pretty busy). Besides whole-home surge protection, now I'll have to
figure out how to insert surge arrestors into the telephone and cable
lines. I checked with my electric company who said that they used to
provide surge arrestors as part of some service plan but do not anymore.
I've been hunting around to see if I can find a surge arrestor anywhere
and haven't found one yet. About the only place I haven't checked so
far is to open up the circuit breaker box to see if it is installed
inside the box instead of inserted into a breaker slot.

 

[w_tom]

1) Your cable provider does not provide a surge protector
because they must provide surge protection. That coaxial
cable is required, even by NEC standards, to be earthed before
it enters the building. That earthing is required for human
safety reasons. That earthing also provides surge
protection. Ask them if their installation conforms to these
requirements:

Outer conductive shield is grounded at the building premises as
close to the point of cable entrance or attachment as practicable.
The grounding conductor shall not be smaller than 14 AWG. It shall
have a current-carrying capacity approximately equal to that of the
outer conductor of the coaxial cable.
The grounding conductor shall be as short as prcticable. In one
and two family dwellings, the grounding conductor shall be as short
as practicable, not to exceed 6.0 meters (20 feet) in length.

2) Your phone company is required by NEC standards to have
installed the protector. If they said none is installed, then
either the responder did not understand what is actually
provided, or responder did not understand the question, or
telco is violating National Electrical Code requirements.
Code is quoted:
From the National Electrical Code Article 800.30A:

A listed primary protector shall be provided on each circuit
run partly or entirely in aerial wire or aerial cable not
confined within the block containing the building served so
as to be exposed to accidental contact with electric light or
power conductor operating at over 300 volts to ground.

Article 800.30B Location.

The primary protector shall be located in, on, or immediately
adjacent to the structure or building served and as close as
practical to the point of entrance.

3) Electric companies would install a 'whole house'
protector behind meter. It would look like an extender ring
behind meter. It would look something like this:
http://www.leaintl.com/pdf/MsaCutsheet.pdf
They would charge something like $5 per month for that
protector. In one year, you a minimally sized 'whole house'
protector could have been paid for almost 1.5 times over.

Electric company would never put anything inside a breaker
box. They would install a 'whole house' protector only where
they have jurisdiction. Inside your breaker box is beyond
their jurisdiction. They can refuse to provide service based
upon what is inside that box. But they do not access or
install equipment inside that box.

Minimally acceptable protectors for residential AC electric
are sold in Home Depot as Intermatic IG1240RC or EG240RC or
Siemens QSA2020.

Some examples of third party phone line protectors:
http://www.one.co.uk/catalogue/telebyte/LightSurgeProtect/22PX.HTM
http://www.keison.co.uk/furse/furse11.htm
http://www.citelprotection.com/english/citel_data_sheets/data_line_protection/B280_B380_B480.PDF
http://www.keison.co.uk/furse/furse08.htm
http://www.oneac.com/pdf/917116c.pdf

 

[David LeBrun]

Ok...about quality units...
One of my systems is using an Enermax 300W, another is Antec 350W.
Yet another is the AOpen branded unit which came in a clone case...at
least on its second or possibly third year of use. Yet another case
is most likely one of the bargain units and I'll probably look to
replace it by swapping in my current Enermax and getting a newer
model. I really like my Enermax unit...well built with good quality
long cables and has been in use well over a year. This Antec unit
seems to be very good also (but its fairly new so can't compare it
much yet). Its voltages are closer to spec than the Enermax but we're
talking
+/-0.2v or so. I like the idea of the dedicated voltage rails on the
newer Antec units...claims that the only load required is on +12v
which would be excellent for bench-testing equipment. Its a toss-up
between these two brands for me.

I've seen other brands advertised: KingWin, Zalman, Thermaltake to
name a few. Also see lots and lots of "no-name" units...when I check
in at the local store I see some 400W units for $30...compared to
similar Enermax for $110 and Antec for $120. What's your brand of
choice?

When I think about what would be in a $30 unit...
- you've got the shell...that can't be avoided...maybe make it thinner
metal but it can't be plastic because of heat...can't afford a lawsuit
when racking in only 30 bucks per

- mobo/drive power lines...can't be avoided either...would be useless
otherwise...could possibly get away with using thinner guage wiring to
cut cost...defintly no gold plated connectors

- single fan w/ sleve bearing...these things are a dime a
dozen...definitely not 2 fans since it would add to cost...not that
there's a spot to put the fan since there aren't any ventalation
holes/slats in these units

- the "guts"...after spending money on all the rest...can't spend too
much here...however since there is no packaging or manual or spec
sheet the savings "could" be applied here

Another thing is that you never get a warranty other than the few days
from the store if it doesn't work. I've trashed enough of these
things I could run a drive power line from my basement to the second
floor.

Dave

 

[w_tom]

Its about what is not in the $30 power supply. For example,
does the computer interfere with radio reception? Another
function that was not installed would be a line filter. Up
front and first provided: no long list of specifications, then
it is probably missing essential functions to increase profit;
customer be damned.

Specs have been seen for Enermax and Antec supplies. Where
are the specifications for that AOpen or, say, a Deer supply?
Specs are provided even though less than 1% of consumers
really know what specs mean. However without those specs, then
the 1% cannot identify the failures of that manufacturers
products to the other 99%.

Tom's Hardware demonstrated another problem with many
discount power supplies:
http://www6.tomshardware.com/howto/02q4/021021/index.html

But then outputs of any power supply can be shorted, power
applied, and nothing destructive happens. No fuse should even
blow. That is even what all minimally acceptable supplies
must do. There is little to nothing visible that determines a
quality supply. Secret ingredients are already in the batter;
not visible to a human eye. Above are just some examples of
how to really identify a minimally acceptable supply. But
first on that list - no specs, then no quality.

 

[David LeBrun]

I just looked around at the stuff in the basement.

There is a thick heavy guage wire running from the main AC switch to
the water main about 6' away. This is the house ground...correct? I
tried to see what was behind the meter looking for a "extender ring"
but since the meter is round I don't know whats part of the actual
meter and whats not.

The telephone box (where all the separate jacks are connected) has 3
wires...2 are for the signal and the third I assume is the ground (old
wiring which does not use all 4 signal wires). The "plate" that its
all hooked up on has a big black knob in the center of it. Don't have
the faintest what it is but it looks ancient...maybe a fuse? The
ground wire (or what ever it is) is also ancient looking and after
digging through the rafters I found it isn't connected to anything.
If this is supposed to be a ground line...could I just hook it up the
AC ground at the water main (which is very close)?

As for the cable...it comes in from the pole at the back of the
property to a plate on the side of the house then continues into the
house (on a separate run of coax). This plate has a wire running into
the ground but I don't know how deep. I also ran a copper ground wire
from the satellite dish to this plate.

These types of things I'm a complete newbie with. Hooking up a new
phone line or cable to another room is one thing but I don't touch the
mains box. Been zapped once and DO NOT want to experience that
sensation again!! I know its not easy without seeing but what do you
think of this setup? Improvements needed?

Dave

 

[David LeBrun]

I guess I should say that I'm in Canada...judging by all the UK links
provided I assume the regulations quoted are also UK and probably
different here.

 

[kony]

Ok...about quality units...
One of my systems is using an Enermax 300W, another is Antec 350W.
Yet another is the AOpen branded unit which came in a clone case...at
least on its second or possibly third year of use. Yet another case
is most likely one of the bargain units and I'll probably look to
replace it by swapping in my current Enermax and getting a newer
model. I really like my Enermax unit...well built with good quality
long cables and has been in use well over a year. This Antec unit
seems to be very good also (but its fairly new so can't compare it
much yet). Its voltages are closer to spec than the Enermax but we're
talking
+/-0.2v or so.

The Antec is a better power supply, due to being more accurately
rated. The Enermax is fine if you consider it at roughly 75% of it's
rating, that is, if you want a ~300W power supply you need a 430W
Enermax. That's the catch with Enermax... they start putting $ into
eye-candy before the inside = label.

I like the idea of the dedicated voltage rails on the
newer Antec units...claims that the only load required is on +12v
which would be excellent for bench-testing equipment. Its a toss-up
between these two brands for me.

It's actually more of a marketing-feature than anything else... in
some circumstances it's actually better to have the ability to draw 3V
& 5V from the same source, like when one or the other rails has
particularly high load. Regardless, the no-load situation merely
means it has internal load resisitors, so in everyday use it runs
slightly hotter and wastes energy. It may somewhat reduce RMA and
allow it to more easily be used on a bench, but minimum load of any
standard power supply is easily met once attached to a *normal*
system.

I've seen other brands advertised: KingWin, Zalman, Thermaltake to
name a few.

Kingwin is similar to Enermax in that the level of eye-candy exceeds
the interior quality, but Kingwin is overall lower quality,
particularly in their lower wattage models.

Thermaltake is about average quality, also known as Highpower or
Enlight, made by Sirtec. They are a fine choice and good value
considering they're lower-cost per wattage, but a notch below an
Antec. Zalman are in the same category, but overpriced for little
more than a quiet fan or more conservative fan throttling (which
shouldn't change cost at all).

AOpen is Sparkle/Fortron relabel, a good value and good construction,
about equal to Enermax in the OEM versions, closer to Antec in the
retail. In general these Sparkle/Fortron power supplies are the best
value (unless something else is on sale) but for a high-end, high
power consumption system I'd recommend the 400W, as it's a significant
step up from 350W, moreso than 350W from 300W.

Also see lots and lots of "no-name" units...when I check
in at the local store I see some 400W units for $30...compared to
similar Enermax for $110 and Antec for $120. What's your brand of
choice?

Most power supplies of decent quality bear the manufacturer's name on
the label. Beware of those that don't. Large volume manufacturers of
power supplies used in OEM and workstations, servers, have high-end
power supplies and components bought at high-volume, manufacturered in
high volume, and lower cost per unit. Delta, Antec, & Sparkle are the
best bets for getting not only a decent power supply, but a fair price
for it (though recently the Antec Truepowers have seen a
disproportionate increase in price compared to the competition). Also
consider "PC Power& Cooling", which are usually overpriced but very
good and may be a great deal when on sale.

When I think about what would be in a $30 unit...
- you've got the shell...that can't be avoided...maybe make it thinner
metal but it can't be plastic because of heat...can't afford a lawsuit
when racking in only 30 bucks per

- mobo/drive power lines...can't be avoided either...would be useless
otherwise...could possibly get away with using thinner guage wiring to
cut cost...defintly no gold plated connectors

Cost cannot be the only deciding factor. There is a large difference
(in favor of Sparkle) when comparing a Sparkle $30 power supply to a
generic. There's a greater difference than merely thinner wires or
metal casing... a few of the other differences were mentioned by
w_tom. Gold plated connectors aren't needed except perhaps in a
marine environment or outdoors, where corrosion occurs at much higher
rate.

- single fan w/ sleve bearing...these things are a dime a
dozen...definitely not 2 fans since it would add to cost...not that
there's a spot to put the fan since there aren't any ventalation
holes/slats in these units

The quietest power supplies often use a sleeve bearing fan for the
slight noise reduction, but it does compromise lifespan... the fan is
often the first part to fail. Don't think junk power supplies only
have one fan though, the low-quality generics may have (and some do)
have two fans... the closer they mimic a decent power supply, the more
likely they can get away with charging similar prices, just a little
less expensive to entice purchase. There are plenty of good
single-fan power supplies, but also many fans labeled as ball-bearing
when it's a 1 ball, 1 sleeve bearing, which IMHO is the worst
possible alternative because an all-sleeve bearing can easily be
relubricated if there isn't a stamped-in-metal fan grill, but relubing
a ball + sleeve bearing fan is less successful because the degraded
sleeve bearing has shed particles that get into the ball-bearing, and
relubing it will cause thinner (than original) lubricant to get into
the ball-bearing, making it very loud. This isn't always the case,
sometimes relubing works, but in general the ball plus sleeve bearing
fans are the shortest lived IF the all-sleeve bearing counterpart is
of equal quality and relubed when necessary.

- the "guts"...after spending money on all the rest...can't spend too
much here...however since there is no packaging or manual or spec
sheet the savings "could" be applied here

Corners are probably cut everywhere possible, from labor cost to
manufacturing to testing, parts, warranty, support, etc.

Another thing is that you never get a warranty other than the few days
from the store if it doesn't work. I've trashed enough of these
things I could run a drive power line from my basement to the second
floor.

Dave

Warranty is nice but realistically I'd rather have a good power supply
with a 30 day warranty than a junk one with 3 years... who covers the
REST of the system failing?


Dave