DSL speed

I haven't calculated whether there would be much practical advantage in
raising my speed to the ceiling. Out of curiosity, I might one day
disconnect my household wiring and try my computer at the phone service
entrance. I guess I should use the splitter to protect the modem from a
ring signal.

I wonder: is it limitations in my wiring and the TELCO wiring that keep
me below the ceiling? In that case, how could they offer DSL several
times faster than my plan?

Another possibility is that the TELCO throttles everybody in my
neighborhood back a little so they have enough bandwidth for us all. Do
they work that way? It occurred to me because the TELCO warns that
there are times they don't have the capacity to sign up an additional
DSL customer in a particular neighborhood.

DSL service runs over copper wires that were not intended to carry
digital information when they were installed (in many neighborhoods,
installed decades ago). That can have an effect on the speed you
attain. And then there is some overhead that comes with TCP/IP and will
prevent you from reaching the full 768. And there can be issues inside
your computer that limit how close you can get to the ceiling of your
DSL data rate. Finally, the data rate is capped at the telco which is
why they can sell different speed levels for different prices.

Have you run Apple's Broadband Tuner on your computer? That might gain
you a little bit.

Is your phone line noisy? Mine was, the Qwest DSL people were useless
when I was having speed and connectivity problems but the POTS service
fixed it by moving my service to a new pair. An old guy named Bob with
a couple of hand tools, a test phone, a ladder and 50 years of
experience with cooper wire fixed my DSL service in 15 minutes- which
was more than the bright young things with their Microsoft certificates
could do. They were clueless.

DSL speeds are limited by something called Shannon's Communication
theory. More noise means less data can be transferred.

DSL also limits the number of subscribers that can share a cable.
You don't share bandwidth with just your neighborhood. You share
bandwidth with everyone who connects to that CO (central office).
Defined are two completely different limitations.

My modem is a ProLine E90-610030-06.
There is a status page, but I don't see an s/n ratio. The speed is
consistent, so it wouldn't be the kind of noise that gets better and
worse, and it's not audible.

That sounds like a possibility. If they have bandwidth for 60
subscribers at full speed but they have 70, they cut everybody back 15%
and it's not much hardship.

Apparently POTS lines can handle 6Mbps if you buy that plan. A lot of
people buy 1.5Mbps and 3Mbps DSL.

I installed it but now regret it.

http://www.macworld.com/news/2005/12/16/brdbndtuner/index.php

Some have had kernel panics after installing it. They recommend
avoiding it unless you have FIOS (622Mbps). My DSL is a thousand times
slower.

With the protector and phone, the s/n is 27 down and 21 up.
Without " " " " 31 22

Speed is unchanged, so I guess noise isn't my limiting factor.

Either way, attenuation is 21.5 down and 11 up.

Now start looking for other items that contribute to reduce S/N
ratio and speed. Once you have found everything, then power cycle the
modem so that it can learn a new speed for a better line. Once modem
takes on a speed that is sufficient, the modem either will tend to
stay at that speed or may drop lower if noise gets worse.

Meanwhile, your S/N ratios look good. Notice a surge protector that
does not even claim to provide protection from typically destructive
surges also significantly diminishes signal strength. Telco has
already installed (earthed) a superior and effective protector where
their wires meet yours.

Next move on to learn about latency using tools such as ping.

Reducing noise 4db had no effect on speed.

What am I supposed to find? I unplugged the phone line to remove the
surge protector and the phone. The modem took a minute or so to
establish DSL and an internet connection.

Attenuation is the same with or without the protector, 21.5 down and 11 up.

The IEEE and the NIST both say a point-of-use protector is worthwhile
for a two-link device.

I ping. Would everybody tied to the local TELCO office have the same
ping times?

Speed numbers meant nothing until after the modem was power cycled.
Your post implied speed measurements taken before power cycling; not
useful information.

Meanwhile, IEEE and NIST are blunt which protector would be
effective: one that is earthed. Your protector that also degrades
signal by 4 dB may even contribute to damage of your DSL equipment.
Telco at the other end of your DSL line also wants their protector
distant from the DSL equipment so as to provide protection. Telco
installs their protectors as defined by IEEE and NIST. Effective
protection means a short connection to earth and distant from
protected transistors. A protector installed free on your phone line
by the telco does both (if properly earthed).

Latency times should be same for all other users of that same CO
equipment connecting to same web sites. If bandwidth consumption by
other users causes slower service, then those ping latency times would
decrease (improve) substantially late at night when other users are
not consuming bandwidth. Increased latency should have no
relationship to your connection between DSL modem and their CO.
Latency change should be a symptom of DSLAM (and things on other side
of) that everyone accesses.

Why do you say that? If a modem is reconnected, doesn't it have to
renegotiate a speed?

That's why they recommend a point-of-use protector. Otherwise, a ground
loop can flow through the equipment.

Why do you say that?

Not around here. The IEEE calls for bonding electrodes. The Electrical
Code calls for it, but the TELCO doesn't believe in it. Without bonding
you'd be better off with no TELCO ground. The TELCO's policy cost my
mother and my neighbor across the street their computers.

That's one advantage to a point-of-use protector: a short path to a
common ground.

Excellent information on surges and surge protection is available from
the IEEE at:
http://omegaps.com/Lightning%20Guide_FINALpublishedversion_May051.pdf
And a guide is also available from the NIST at:
http://www.nist.gov/public_affairs/practiceguides/surgesfnl.pdf

The IEEE guide explains plug-in suppressors work by CLAMPING the
voltage on all wires (signal and power) to the common ground at the
suppressor. Plug-in suppressors do not work primarily by earthing.
The IEEE guide explains connection to earth occurs elsewhere. (Read
the guide starting pdf page 40).

Both the IEEE and NIST guides say plug-in suppressors are effective.

Note that all interconnected equipment needs to be connected to the
same plug-in suppressor, or interconnecting wires need to go through
the suppressor. External connections, like phone, also need to go
through the suppressor. Connecting all wiring through the suppressor
prevents damaging voltages between power and signal wires. These
multiport suppressors are described in both guides. This is the same
protection for "two-link" devices E Z described.

Effective protection also requires a "single point ground" - a short
connection from phone, CATV, ... entry protectors to the earth
connection wire at the power service. Connections have to be short to
prevent a surge earth current on the connecting wire from producing a
large voltage difference between the signal and power wires. Short
connection between systems is more important than short connection to
earth electrode.

--
bud--

As demonstrated below on Page 42 Figure 8, it is the 'point of use'
protector that creates a ground loop 8000 volts destructively through
an adjacent TV. Effective protectors earth before surges enter the
building. Once inside, the build is chock full of potential ground
loops destructively through appliances.

EZ Peaces - you did not even know basic electrical concepts such as
the significance of 'signal to noise' ratio. Suddenly a plug-in
protector does magic - somehow stops what 3 miles of sky could not -
does what even its manufacturer will not claim? Why? It even
degrades your DSL signal by 4 dB - an unacceptable degradation. They
forgot to mention that degradation also? What else did that power
strip manufacturer forget to mention?

Why do you believe urban myths while ignoring manufacturer spec
numbers? Did you notice the manufacture will not even claim what you
have posted?

Why do you think your telco installed the same type protector used
to protect their $multi-million switching computer? Are they fools as
you seem to believe? Why do you think they don't use plug-in
protectors? Why do you think their protector has a dedicated earthing
wire that must be short? Why do you think both NIST and IEEE state
bluntly that protectors are connecting devices (divert) to earth? In
each case, an earthed protector provides effective protection.

Even Bud's sources demonstrate plug-in protectors as ineffective.
Once Bud cited Martzloff repeatedly. But Martzloff also said - a
first point in his 1996 IEEE paper conclusion - that plug-in
protectors can even contribute to electronic damage. In direct
contradiction to what E Z Peaces has posted:
> Conclusion:
> 1) Quantitative measurements in the Upside-Down house clearly
> show objectionable difference in reference voltages. These occur
> even when or perhaps because, surge protective devices are
> present at the point of connection of appliances.

Bud demonstrate the problem - a ground loop destrutively through
adjacent electronics:
http://omegaps.com/Lightning%20Guide_FINALpublishedversion_May051.pdf
Page 42 Figure 8 shows a plug-in protector destroying an adjacent TV
with 8000 volts. Protector was too close to TV and too far from
earthing - had all but no earth ground. 8000 volts destructively
through a TV diverted destructively via a plug-in protector.

That same protector that degrades DSL signal by 4 dB may also do
same damage to the DSL modem, or router, or computer.... Effective
protector earths before a surge can enter the building. No surge in
the building means no ground loops. No wonder Martzloff also said the
plug-in protector can contribute to appliance damage.

No earth ground means no effective protection - which is why the
telco protector is earthed. Even a plug-in protector manufacturer
does not deny it. They also forget to mention signal degradation to
both DSL and cable signals. They forget to mention 8000 volts
destructively through a TV when a protector is not properly earthed.

Why does your telco not use plug-in protectors? Just like in IEEE
and NIST papers, telco papers also define earthing as essential to
protection. Same type protector that protects their $multi-million
switch computer is also properly earthed at subscriber end. How do
we know it provides protection? It has the essential 'less than 10
foot' connection to earth (assuming the homeowner provided that earth
ground). Exactly what the NIST demands on page 6 (Adobe page 8 of 24)
in Bud's other citation:
http://www.nist.gov/public_affairs/practiceguides/surgesfnl.pdf
> You cannot really suppress a surge altogether, nor
> "arrest" it. What these protective devices do is
> neither suppress nor arrest a surge, but simply
> divert it to ground, where it can do no harm. So
> a name that makes sense would be "surge diverter"
> but it was not picked. So, for the rest of this
> booklet, we will stick to the most popular "surge
> protector".

E Z Peaces - even your protector manufacture will not claim what you
have posted. Your plug-in protector has degraded DSL signals by 4 dB
- unacceptable. As Bud has so politely demonstrates - that protector
may even earth surges destructively through your DSL modem. Did you
really believe it would stop or absorb what three mile of sky could
not. That is the junk science assumption. Did you really believe it
will stop what three miles of sky could not?

An effective protector earths before surges can enter a building.
Once inside that building, surges will find numerous and potentially
destructive paths through household appliances - ground loops. But
then earthing is what all reliable facilities install. They don't use
plug-in protectors that would also degrade a DSL signal by 4 dB.

According to the diagram, that surge came /from/ the /earth/. The
second TV was at risk because it was not plugged into the point-of-use
protector.

What on /earth/ makes you say that?

Unacceptable to whom? As it doesn't affect my speed, I find it acceptable.

What's that?

Why do /you/ think that's the case?

Why do you think they recommend point-of-use protectors?

The book says Figure 8 shows how an improperly used protector may not
fully protect. Figure 7 shows how a properly used protector protects.

Most surges come through the grounds. How will clamping power lines at
the entrance prevent that?

Two years ago lightning hit my house. It damaged two chimneys, knocked
a hole in the roof, and blew off siding. It got my air conditioning and
a CD player. It got stereos in three rooms. (They used only
rabbit-ears antennae.) One stereo wasn't even plugged in.

One of the destroyed stereos had been turned off, sitting ten feet from
my computer. I was online. My computer froze with a rainbow across the
screen. I restarted and went back online. There was never any sign of
damage to my computer equipment. So if you want to tell everyone my
point-of-use protector stopped a lightning bolt, I won't quibble.

I lost a computer in 1994 because I depended on a whole-house protector
and the grounds installed by the power company and the TELCO. I made
improvements. In 1998, I was online when lightning hit a tree 30 feet
from my service entrance. The TELCO had to send a man to climb the pole
and replace a fuse, but I had no damage.

In 2002 lightning lit up the neighborhood in orange when it hit a tree
65 feet from my service entrance. I was online but had no damage.

When my house was hit in 2005, my computer and phone equipment were
fine. My neighbor across the street wasn't so lucky. The ground surge
from the strike on my house wiped out everything tied to his phone line:
a computer, two cordless phones, and a satellite TV system. Like you,
he had depended on the protector supplied by the TELCO. For years I'd
been warning him that the TELCO's installation was not up to code.

After the strike, my neighbor called the TELCO, who sent a man, who
looked at the installation and said it was grounded properly and had not
contributed to the damage. My neighbor told me I was wrong. Well, his
BIL is an executive with the power company. He backed me up. The TELCO
installation was not up to code and had caused the damage. I'm not
calling the TELCO fools, but I would not count on them for lightning
protection.

> Even Bud's sources demonstrate plug-in protectors as ineffective.

w_ is apparently halucinating.

w_ forgets to mention that Martzloff said in the same 1994 (not 1996)
document:
"Mitigation of the threat can take many forms. One solution.
illustrated in this paper, is the insertion of a properly designed
surge reference equalizer [multiport plug-in surge suppressor]."

In 2001 Martzloff wrote the NIST guide that says plug-in suppressors
are effective.
In direct

Lacking technical arguments w_ has to lie about sources that
contradict his belief in earthing.
The illustration in has a surge coming in on a CATV drop. There are 2
TVs, one is on a plug-in suppressor. The plug-in suppressor protects
the TV connected to it. It does *not* contribute to damage of the
second TV. Without the suppressor, the voltage at the 2nd TV is
10,000V. With the suppressor the voltage is 8,000V. The point of the
illustration, for the IEEE and anyone but w_, is "to protect TV2, a
second multiport protector located at TV2 is required". Read the
source.

w_ has a religious belief (immune from challenge) that surge
protection must use earthing. Thus in his view plug-in suppressors
(which are not well earthed) can not possibly work. The IEEE guide

suppressors do not work primarily by earthing. The guide explains
earthing occurs elsewhere. (Read the guide starting pdf page 40).

What does the NIST guide actually say about plug-in suppressors?
They are "the easiest solution".
And:
"Q - Will a surge protector installed at the service entrance be
sufficient for the whole house?
A - There are two answers to than question: Yes for one-link
appliances, No for two-link appliances [equipment connected to power
AND phone or CATV or....]. Since most homes today have some kind of
two-link appliances, the prudent answer to the question would be NO -
but that does not mean that a surge protector installed at the service
entrance is useless."

The question is not earthing - everyone is for it. The only question
is whether plug-in suppressors work.

There are 98,615,938 other web sites, including 13,843,032 by
lunatics, and w_ can't find another lunatic that says plug-in
suppressors are NOT effective. All you have is w_'s opinions based on
his religious belief in earthing.

But both the IEEE and NIST guides say plug-in suppressors are
effective. Read the sources.

And never explained:
- Why do the only 2 examples of protection in the IEEE guide use plug-
in suppressors.
- Why does the NIST guide says plug-in suppressors are "the easiest
solution".
- Why did Martzloff say in his paper "One solution. illustrated in
this paper, is the insertion of a properly designed surge reference
equalizer [multiport plug-in surge suppressor]."

Bizarre claim - plug-in surge suppressors don't work
Never any sources that say plug-in suppressors are NOT effective.
Twists opposing sources to say the opposite of what they really say.
w_ is a purveyor of junk science.

--
bud--

It came from earth because the 'whole house' protector was not
properly installed. Instead, the homeowner used a plug-in protectors
and suffered 8000 volts destructively through the TV.

Electronics already have internal protection. Anything that plug-in
protector was going to accomplish is already inside appliances. But
internal appliance protection is dependent on a 'whole house'
protector properly installed. Why? Earthing - not the protector -
defines protection - even on Page 42 Figure 8.

E Z Peaces examples apparently demonstrate the same thing. Also how
the telco protects their $multi-million switching computer that
connects to overhead wires all over town. Also how commercial
broadcasters install their protection from so many direct lightning
strikes.

For example, Orange County FL was suffering lightning damage in
their emergency facilities. They fixed the problem. Did they install
any plug-in protectors? Of course not. Orange County needed real
world protection. They fixed the earthing:
http://www.psihq.com/AllCopper.htm

Surges come from overhead or from ground. A nearby struck tree may
actually be a direct strike to appliances in the building. Why?
Improper earthing. I have doubts about your claims that most
destructive surges enter via the ground. Even wires down the street
struck by lighting are a direct strike to household appliances. But
reason for a properly earthed 'whole house' protector means a system
that provides both equipotential and conductivity. Both are required
so that all sources of surges are made irrelevant.

So what does the plug-in protector claim? It avoids the entire
discussion. It has no earthing. It somehow will stop what three
miles of sky could not.

Let's assume the multiport protector only intends to protect DSL
using equipotential - as Bud once claimed. Bud once made those
claims. Well AC electric and the incoming phone line connect to that
protector. What about the wire outgoing to computer or routers? If
every one of those eight wires in every cable is not part of
protection, then equipotential is not achieved. That's one port not
included in a multiport protector meaning damage can occur.

Another port is the surface that modem is sitting on. Another port
not included in the multiport protector again means equipotential is
not achieved.

For a protector to provide equipotential to the adjacent TV, then
every wire in every cable and every mechanical support must be
integrated in the protector circuitry. That does exist inside the
appliance. But a plug-in protector on your DSL modem does not.

Why does the US Air Force instead demand 'whole house' type
protection? Equipotential is achieves beneath the entire building.
Earth beneath the building means every port in every appliance has
multiport protection. BUT conductivity also must be provided.
Therefore the 'whole house' protector also has the mode conductive
path to earth.

Plug-in protectors do not provide equipotential beneath a building
AND avoids all claims of conductivity since effective earthing does
not exist. That is why Page 42 Figure 8 - the miracle of a plug-in
protector - in reality earths a surge destructively through an
adjacent TV.

All electronics already contain effective protection. Protection
that assumes surges - from both overhead and underground - will be
earthed by the 'whole house' protector with single point earth
ground. No earth ground means no effective protection. Plug-in
protectors clearly have no earth ground which is why Page 42 Figure 8
shunts thousands of volts into both TVs.

Why do Polyphaser application notes discuss earthing extensively?
http://www.polyphaser.com/technical_notes.aspx
Polyphaser is an industry benchmark. Earthing defines the quality of
protection - not some silly plug-in solution that is also typically
undersized. But it is so profitable.

Polyphaser even discusses a surge entering underground to cause
electronics damage. What is the solution? Single point earthing in a
'whole house' protector solution:
http://tinyurl.com/38v2dv
> 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.

Why is the TV destroyed by 8000 volts? Plug-in protector was too
close to electronics and too far from a single point earth ground.
Why did that surge exist? Missing - earth ground - is what Bud and
plug-in protector manufactures avoid claiming a plug-in protector is
the miracle solution. A protector is only as effective as its earth
ground - as noted by so many above and responsible sources, by Page 42
Figure 8, and even implied by E Z Peaces.

That protector on a DSL modem does not cover all ports. Therefore
it does not provide equipotential as Bud claims. It may even
contribute to modem damage.

Protector does cause an unacceptable 4 dB signal reduction. A
number that would probably be higher if using the faster DSL. But
your DSL signal is so strong that S/N ratio should not account for the
reduced DSL speed. Of course, that must be confirmed by first moving
all devices that reduce dB signal (ie protector), and the power
cycling the modem. Latency would be another number that might report
useful information.

In the diagram, a bonding wire connects the cable grounding electrode to
the power grounding electrode. How would a proper installation of a
whole-house protector change that?
>

Years ago, my BIL's farmhouse would blow light bulbs in thunderstorms,
and electricity would pop from spot to spot in the house. He had a
whole-house protector. His power and phone service were grounded to the
same electrode with short wires. I found all the places where the house
was grounded by supply or drain plumbing. I bonded those pipes to the
main electrode. That stopped the snapping and blown bulbs.

Then he got a modem. Several times a year he had to have it repaired
under warranty after thunderstorms. He tried a phone-line surge
protector a few feet from the modem. It made no difference.

It seemed that the manufacturer would rather keep repairing the modem
than discuss lightning, but finally my BIL got an electrical engineer at
the company to talk turkey. The engineer told him to use the same
point-of-use protector for his power and phone lines. That was nearly
20 years ago and he hasn't had any trouble since.

The first problem with his original protector was that it was the
gas-tube type. TELCOs like to use them at service entrances because
they are durable, but the clamping voltage was too high to protect his
modem, and it may have been too slow. The second problem was that it
was grounded to a different outlet from the one powering his computer
and modem. Surges from lightning rise and fall so fast that the grounds
at two outlets could differ by hundreds of volts due to inductive reactance.

The combination surge protector solved my BIL's problem by clamping his
phone line and power line to the same ground point close to the
equipment. That sounds like what polyphaser says in recommending
bringing the ground into the telephone room.

I'm interested in that. In the past I read up on 56k and it's a
mathematical marvel. How does DSL work? When the modem is idling, the
ethernet and internet blink only a little. When I'm downloading, they
cycle off and on several times a second. I wonder if that means the
data is coming in pulses that are faster than 768k.

Effective protection requires two things. First, it must be the
most conductive connection to earth. When transient currents (current
is the defining parameter) are short but tens of thousands of amps,
then the best conducting device is the path those currents will take.
When current is not properly conducted, then high and destructive
voltages occur. There is not 'million volts of damage' in lightning
when is properly conducted to earth.

For most homes, that means at least a 10 foot earth ground rod.
Soil should be conductive (not granular like sand). Rod should be in
earth that is below the frost line and best when always moist. Since
these conditions don't always exist, then we use other solutions such
as a network of ground rods with buried internconnecting wires, a halo
ground, Ufer ground (installed so that lightning would not even cause
munitions explosions), a ground plate, etc. Notice that the water
pipe is not listed as a best ground. Best earth ground is a connection
to earth that is most conductive. This sometimes made challenging
when building also uses a well pump.

Second, effective protection provides equipotential. Since we
cannot provide sufficient conductivity, then we also must provide
equipotential. That means earth beneath all parts of the building is
predominately at a common voltage. Some achieve this by a deep
grounding rod. Other better solutions include a halo or Ufer ground
that loops the building. Latter solutions are more often installed in
sandy locations as demonstrated by pictures from Gfretwell:
http://members.aol.com/gfretwell/ufer.jpg
Others also discuss Ufer solutions:
http://www.psihq.com/iread/ufergrnd.htm
http://scott-inc.com/html/ufer.htm

In each case, both equipotential and conductivity are increased.
Again, a protector being only as effective as its earthing.

For example, lightning strikes a nearby tree. The circuit is a
connection from cloud, down the tree, many miles through earth, to
earthborne charges. But a shorter path is up a cow's hind legs and
down his fore legs. Cow is electrocuted because he did not have a
single point ground.

Number of ways to achieve that single point ground. Have cow stand
only on its hind legs while keeping them together. Another would be a
halo ground wire buried around the cow. Electricity that does not
have both an 'incoming and outgoing' path or tht finds a better path
therefore not pass through the cow. BTW, same reason why humans are
advised to keep feet together if caught out during thunderstorms -
create a single point ground.

Humans are also safer inside buildings because wood is a conductor.

Those two points are the principles. What defines effectiveness of
a protector? That single point earth ground. Therefore all utilities
must enter a building at a same point. The most conductive path to
earth being earth connected to each wire in each cable by 'less than
10 feet'.

Another important concept - impedance. For example, wall receptacle
may be 50 feet from breaker box - even longer to earth ground. But
wall receptacle is only a safety ground; not earth ground. Wire
resistance may be less than 0.2 ohms. But wire impedance to a surge
may be 120 ohms. A trivial 100 amp surge would leave wall receptacle
at 100 x 120 or something less than 12,000 volts during a surge. If
trying to earth via that wall receptacle, well, a room is chock full
of other conductive paths. Surge currents will go elsewhere. That is
the point of Page 42 Figure 8 - the 8000 volt damaged TV.

Demonstrated is why surge currents permitted inside a house find
numerous and destructive paths inside that house. Demonstrates also
why a connection from each utility wire to single point earth ground
must be 'less than 10 feet' ; or shorter. Electrically shorter also
means no sharp bends, no splices, not inside metallic conduit, etc.
Protection means earthing wire also must be separated (at least one
foot) from other non-grounding wires. AC recpetacle ground violates
every requirement - why a plug-in protector has no earthing.

If breaker box is earthed by a bare 6 AWG ground wire up over
foundation and down to earth (as is so common because wood is simpler
to drill), then that panel is not sufficiently earthed. That ground
wire must go through foundation and down to an earthing electrode.
Shorter distance. Less bends. Earthing wire also be separated from
other wires.

Of course, every earthing wire must make that short connection to
same electrode. Also necessary for equipotential - so that all wires
entering a building are at a common voltage potential during the
entire surge. We call this 'single point earth ground'. A single
ground rod, a network of rods, or a halo ground, still, each utilities
should have separate and dedictated earthing wire so that all meet at
a same point on that earthing electrode.

However since even electricians don't understand these concepts,
then some homes are constructed with utilities entering in multiple
directions. That created the above 'dead cow' situation. One utility
offers a solution:
http://www.cinergy.com/surge/ttip08.htm
With an interconnecting wire buried, then every utility is making some
kind of connection to a common ground. Clearly, better is for all
grounds connect at a same point. Also better would be a wire
encircling the building - halo ground. Bottom line conclusion remains
same. Earthing defines the quality of that protection system.

So many professional discuss these concepts. A days worth of
reading is found in two posts in can.internet.highspeed on 22 Jun
2007 and 28 Jun entitled "Of lottery tickets and lightning" in:
http://tinyurl.com/32v3le

Note, for example, a picture in an app note from an industry
professional:
http://www.erico.com/public/library/fep/technotes/tncr002.pdf
Two structures. Each has its own single point earthing. All wires
(even underground) that enter a building must first make that single
point ground connection. Wire from tower to building must connect to
both structure's single point ground before entering structure. To
make both earthings even better, a buried ground wire interconnects
both single point grounds.

Enhancements detailed above are to post-1990 National Electrical
Code requirements.

More on that earthing connection. Every incoming wire in every
cable must make that 'less than 10 foot' connection to single point
earthing system. That connection is made by hard wire (cable TV and
satellite dish) or made via a 'whole house' protector (AC electric and
telephone). The protector was not protection. The protector acts
like a switch - connecting each wire to earth only during a surge.
What is the most important parameter of that connection? First - it
must be as short as possible (again - low impedance means shorter wire
length).

Polyphaser makes a protector that has no connection to earth. That
protector mounts ON earth ground - zero foot connection. That short
distance (and no sharp bends, etc) means single point ground provides
better conductivity and equipotential.

Since a protector is a switching device, then it has limitations.
Gas Discharge Tubes were some of the best conductors during surges - a
standard solution in the early 20th Century so that direct lightning
strikes would not cause damage. But surges tended to cause GDT
electrodes to vaporize - contaminate the gas - cause its threshold or
let-through voltage to increase. GDTs were obsoleted by carbons as
demonstrated in these pictures:
http://www.inwap.com/inwap/chez/Phoneline.jpg
http://dougk-ff7.net/DSCN0529.jpg

Carbons (lower capacitance MOV like devices) where then obsoleted by
semiconductor devices now used today. Some examples in residences:
http://www.alarmsuperstore.com/bw/bw%20connectors.htm
http://www.basshome.com/product_3676_detailed.htm
http://www.basshome.com/product_4680_detailed.htm
http://www.citynet.net/supportdp.cfm?article=8
http://www.inetdaemon.com/tutorials/telecom/pstn/nid.shtml
For communication protectors, low capacitance is essential. Phone
system was installed 50 years ago with equipment that can still
support higher frequencies in DSL. Low capacitance is why cable
companies also recommend no protector on their digital cable service.

Cable company needs no protector. For better protection with lower
capacitance, the earthing connection is made with a $2 ground block
and 12 AWG wire. Again, that connection must be short which is why a
cable TV ground to a water faucet or AC electrical receptacle is not
earthed. Check your cable. Did the installed ground 'less than 10
feet' direct to the electrode? If not, then even DSL is at risk.

With cable companies now providing service that must be 'phone'
reliable, suddenly cable installers were taught how to earth. Why
they are earthing is not taught to installers or electricians. Those
reasons are provided here.

Again, since a protector is a switching device, then it has
limitations. But AC electric protectors need not have low
capacitance. Therefore MOVs are a preferred technology. Since the
protector is nothing more than a switch, then that 'switch' has a life
expectancy. Life expectancy is defined in charts that relate current,
length of that transient, and number of transients. Ballpark
measurement for charts with superior curves is called joules. For
residential 'whole house' protectors, minimum is about 1000.joules and
50,000 amps. That number is increased where number of surges,
frequency of surges, or even better protection during each surge is
desired.

Minimal 'whole house' protector sells for less than $50 in Lowes and
Home Depot. Even better units - more joules - are sold there or in
electrical supply houses.

Average lightning strike is 20,000 amps. That strike current is
shared by each household and utility installed protection. Therefore
any minimally required protector is more than sufficient to earth
surges and remain functional. IOW the human should never even know a
surge existed. That is not the case with so many plug-in protectors
that are grossly undersized. Therefore the naive will claim "my
protector sacrificed itself to save my computer". They only assume.
In reality, surge current hit both computer and protector
simultaneously - protector located to close to computer. A surge too
small to overwhelm protection inside a computer instead vaporized a
grossly undersized plug-in protector.

That unacceptable failure then promotes sales among the naive. To be
electrically equal to a 'whole house' protector, the plug-in protector
must be 3000 joules or higher. No wonder current technology
protectors create these scary pictures:
http://www.hanford.gov/rl/?page=556&parent=554
http://www.westwhitelandfire.com/Articles/Surge%20Protectors.pdf
http://www.ddxg.net/old/surge_protectors.htm
http://www.zerosurge.com/HTML/movs.html

Effective protectors remain functional after each surge - do not
threaten human life. Ineffective protector will avoid all discussion
about earthing when earthing is THE most critical component of an
effective protection 'system'. Protector being only as effective as
its earth ground. The protector being sufficiently sized so as to
always be functional for all surges.

How common are surges? Average is about one surge every seven
years. Obviously that number varies significantly in different
regions. Number can also vary significantly within the town. Geology
is a major factor here. For example, a house at the end of an
electrical distribution system may be in the best path to earth
lightning that frequently strikes utility wires down the street.
Therefore that last house suffers more damage. Some areas are located
above a massive igneous deposit making that area a more preferred
strike area. Installation of a transcontinental pipeline can make the
ground a better conduit for lightning. Some have reported better
protection from nearby tall pine trees. Even one region with numerous
graphite veins made equipotential difficult to achieve without halo or
Ufer grounds.

But again, what defined the effectiveness of protection? Earthing.
Lightning seeks earth ground. It will find that path destructively if
permitted inside a building. Since 1970, all buildings should have
been constructed as if the transistor existed. Today, we still do not
do so. Gfretwell's pictures and other citations above demonstrated
Ufer examples for transistor protection. Bill Otten demonstrates
solutions in:
http://home1.gte.net/res0958z/
Those Orange County solutions were posted previously. Bell System
Technical Journal discussed these concepts when telephone switches
were being converted to transistors in 1960. Qwest now wants prefers
halo ground to Ufer grounds. But in every case, they are defining the
most important component in a protection system. So many others
discuss earthing in a days worth of reading at:
http://tinyurl.com/32v3le

In each case, the one component required in every protection
'system' is an earthing electrode. Protectors do not provide that
protection. Protectors are nothing more than connecting devices to
protection. That protection is the quality of and connections to
earth ground. Earthing is what defines both conductivity and
equipotential - the all so crtical sigle point earth ground.

Discussed above is only 'secondary' protection. Also inspect your
'primary' protection system installed by utilities:
http://www.tvtower.com/fpl.html
Strikes to household appliances are often direct strikes to wires down
the street. According to IEEE papers, the majority of that strike
current (maybe 60%) is earthed by that primary protection system. But
if those ground wires are disconnected (as First Energy was ignoring
to cut costs and to pay for their acquisitions - a bean counter
mentality), then your home is at significantly higher risk to surge
damage. But again, what defines the protection?

Each protection layer is defined by one thing: the layer's earth
ground. A plug-in protector can only be as good as its connection to
the same earth ground used by a 'whole house' protector. Why would a
surge via a plug-in protector use that same ground when it did not
take a shorter path through a 'whole house' protector?

How to make that plug-in protector more effective? Cut its power
cord shorter and move it to a receptacle that is closest to breaker
box earth ground. Or instead, use that same money to buy a protector
with more joules and a dedicated earthing wire. A 'whole house'
protector is tens (maybe 100) times less money per protected
appliance. And with a better earthing connection, it does what the
plug-in protector does not even claim to accomplish (see its numeric
specs).

Described are how both equipotential and conductivity provide
protection AND how more is accomplished for so much less money. Of
course, if these solutions were installed before footings were poured,
then a superior solution costs even less. But we still don't build
homes as if the transistor exists. In part because power strip
protector promoters have done so good at perverting knowledge - get so
many to believe a plug-in soluton will solve all. The protector is
only as effective as its earth ground - as so many above professionals
note. As even Ben Franklin demonstrated in 1752.

Notice how Bud will even resort to insult to protect that massive
profit margin on protectors that don't even claim to protect from the
typically destructive type of surge. Notice that he follows me
everywhere trolling to confuse others. Plug-in protector
manufacturers have him to promoted confusion and myths. Their grossly
overpriced products don't have that dedicated earthing wire; may even
contribute to electronics damage as demonstrated on Page 42 Figure 8.
A protector too close to electronics (as much as 12,000 volts between
AC wall receptacle and earth) and too far from earth ground ('less
than 10 feet is required) simple earths a surge 8000 volts
destructively through the adjacent TV.

Finally, how is the POTS modem typically damaged? Telephone line
already has an earthed 'whole house' protector. Incoming on AC
electric, through motherboard (this path made easier by an adjacent
power strip protector), through modem, then to earth via phone line.
First current follows in everything in that path. Then something
fails. A most common failed part in modems is the PNP transistor that
drives its off-hook relay. Since that relay was typically only rated
for 500 volts, then coil was electrically connected to wiper and
phone, destructively via the PNP transistor. Repaired so many modems
by simply tracing the surge path to find the damaged part. One
lightning damaged modem here has since works just fine for more than
seven years. Tracing the damage path is just another way we learned
this stuff.

How to eliminate the modem damage? Properly earth an AC mains
'whole house' protector. AC electric being the most common incoming
path for surges through modems to earth ground.

On Jul 3, 5:52 pm, E Z Peaces <c...@invalid.invalid> wrote:

> ...

> ...
>
> The combinationsurgeprotector solved my BIL's problem by clamping his

> ...

Modems are limited not by copper wire (as myths once promoted) but
by how the telephone switching computer (circuit switched technology)
operates. By sampling at a fixed speed, a modem's data rate cannot
exceed the COs sampling rate. And since S/N ratios exist (due to D/A
converters, etc), then 56K modems cannot even achieve 56K.

DSL was pioneered by British Telephone in 1981 - and kept out of
this nation until Clinton pushed through the 1996 Federal
Communication Act. Unlike AT&T, the Baby Bells did not completely
fear new technologies. They just feared to pioneer it. That law all
but required them to innovate.

DSL is packet switched technology - what Isenberg calls the 'Dumb
Network' (Find the paper he cannot post but is found via his web site
at www.isen.com). That means the phone line now connects to two
completely different machines - the Central Office (CO) switch and a
DSLAM. CO switch works only at audio frequencies. DSL operates at
radio frequencies. When 1950 telco equipment was installed with low
capacitance devices and twisted pair wires, then the network would
also support radio frequencies - if the central office switch was
eliminated.

DSL bandwidth is determined by data transmitted in many frequencies
that are, BTW, also the most destructive frequencies in lightning.
POTS (Plain Old Telephone) equipment would 'eat' these frequencies.
So we install filters to block radio frequencies from your phone and
fax machine AND to that CO switching computer. Wire from your DSL
modem to their DSLAM must be nothing more than wire - so that radio
frequencies have a direct hardwire connection. All protectors must be
low capacitance type.

Your DSL is a 768K version. Therefore it uses lower frequencies and
a distance limit accordingly. Your S/N ratio is quite good. If using
higher bandwidth DSL, then even higher radio frequencies are used.
Anything that would diminish your DSL signal (ie capacitance in that
power strip protector) would diminish a faster DSL signal even more.
Of course, that higher data bandwidth also means wire connection to CO
has a shorter maximum distance - signal attenuates more with longer
wire.

Another limitation is the number of DSL subscribers within one
multiwire telephone cable. I don't remember the limits. But cross
talk limits the number of DSL users to a speculated 60% of wire
pairs. Audio cross talk between twisted pair cables is not
significant. But radio frequency cross talk creates limitations.

Of course the biggest limitation, as usual, was not technical.
Management that feared to pioneer innovation stuck us with POTS modems
when DSL was demonstrated in 1981. Fear of technology is widespread
among companies with an east coast mentality. Thank goodness we
separated AT&T from the Baby Bells so as to make innovation possible.
AT&T management being some the most anti-innovation minds in the
nation - such as Robert Allen whose MBA was from Harvard.

http://www.linktionary.com/d/dsl.html
This site says ADSL uses 300-700kHz upstream and 1-10MHz and above
upstream. I wonder why my down side shows a little more attenuation.

I wonder what kind of modulation DSL uses. 56k uses phase and amplitude.

When they advertise 768kb/s, I wonder how many bits in a byte in a DSL
transmission. I've read that with dialup, each eight-bit is packed with
two more bits and sent as a ten-bit byte. So if you had a perfect
connection, about 53kb/s, your computer would receive the file at 42kb/s.

Water pipe is not listed because w_ thinks it should never be used as a
grounding electrode, even though it has been required to be used for a
very long time, and for a urban metal water system is the most
conduuctive to earth of any electrode available at a house.

From
http://www.lightningsafety.com/nlsi_lhm/grounding_definitions.html
“Halo Grounded Ring: A grounded No. 2 wire, installed around all four
walls inside a small building, at an elevation of approx. six inches
below the ceiling. They are used around transmitter equipment.
How would a halo ground help?

Misuse of halo again.

A Ufer/concrete encased electrode may give you equipotential depending
on how much the rebar is tied together.

With a ground rod as a rule of thumb 70% of the voltage drop from a
ground rod is in the first 3 feet. Assume the power system is earthed
with only a ground rod. If you have a very good rod-to-earth resistance
of 10 ohms and a modest 1,000A surge earth current, the voltage from the
power ground bar to `absolute' earth is 10,000V. From the power/signal
ground system to earth beyond 3 feet will be 7,000V or more. This can
show up, for instance, in basements, like a E Z’s washing machine to a
concrete floor. Or as the IEEE guide notes, at outside pad mounted
compressor/condenser units.

You are not likely to get equipotential around a house.

The religious belief (immune from challenge) that surge protection must
use earthing. The IEEE guide explains plug-in suppressors work primarily

ground at the suppressor, not earthing. The guide explains earthing

Francois Martzloff, who wrote the NIST gide, has written “the impedance
of the grounding system to ‘true earth’ is far less important than the
integrity of the bonding of the various parts of the grounding system.”
The point of a "single point ground" is that entrance protectors for
CATV, phone, ... be connected with a *short* wire to the conductor to
the earth electrode at the power panel. With a large surge there will
always be a difference from the house ground to ‘absolute’ ground. The
goal is for the power and CATV and phone 'grounds' to rise together.

The point of Fig.8, for anyone who can read and think, is "to protect
TV2, a second multiport protector located at TV2 is required". The
failure behind the illustration is lack of a “single point ground” - the
CATV entry point is too far from the power service. That is the
condition at a significant percentage of houses. The IEEE says that “if
the CATV, satellite, or phone cables do not enter the building near the
service entrance, the only effective way of protecting the equipment is
to use a multiport protector.”

As E Z has noted, a service panel suppressor will provide absolutely NO
protection to either TV is this case.

w_ finds it impossible to understand the IEEE illustration because it
conflicts with his religious belief in earthing.

And there will be arc-over in panels or receptacles at about 6,000V -
you won’t get 12,000V between H-N-G.

Needs no protector? The IEEE guide notes that the voltage between cable
center conductor and sheath is limited by the breakdown of F-connectors
which is typically 2-4,000V. The guide notes that connected equipment
can be damaged at those voltages. Plug-in suppressors are likely to
clamp the voltage to a reasonable level.

w_ denies his own hanford link. It talks about "some older model" power
strips and says overheating was fixed with a revision to UL1449. That
was 1998.

It is a lie that “current technology protectors create these scary
pictures.”

But with no valid technical arguments all w_ has is pathetic scare tactics.

The religious belief in earthing #2. The IEEE explains plug–in
suppressors work primarily by CLAMPING, not earthing.

"To quote w_: 'It is an old political trick. When facts cannot be
challenged technically, then attack the messenger."

I resort to insults - w_ just called me a stodge for manufacturers.

Being evangelical in his belief in earthing, w_ trolls google groups
for "surge" to paste in his religious tract to convert the heathens.
Unfortunately some, like E Z, remain pagans.

w_’s latest epistle is sometimes very good, sometimes garbage. But
almost entirely irrelevant to plug–in suppressors. The question is not

w_ has still not found a link to another lunatic that says plug-in

Still never explained:

Twists opposing sources to say the opposite of what they really say -
hanford, IEEE guide.
Attempts to discredit opponents.
w_ is still a purveyor of junk science.

--
bud--

Traditional asynchronous serial data streams had eight data bits,
one start bit and one parity bit. (We are ignoring other, less
popular, serial protocols.) Ten bit data stream delivers eight bits
of data. So yes, 52K bit transfer rate would result in about 42 bit
of data. And then it gets more complex. The data may be compressed
which is why a 52K data stream might deliver 5K byte (word or
character) transfers.

IP packets also have overhead. Actual data transfer may be faster
if packet sizes are larger. However when measuring data transfer,
well packets are asynchronous. Bit rate can be very high. But
latency between those data packets can result in a low data rates.

Attenuation also is not entirely clear. For example, your AM radio
must receive the station 50 miles away just like the same station was
only 1000 feet away. So radios have automatic gain control (AGC).
Closer (stronger) stations are attenuated so that the signal does not
saturate amplifiers.

Is that what attenuation is measuring on your DSL modem? It's not
always clear since many modems short us on facts. If attenuation is a
measure for incoming (downstream) signals, then what is the upstream
attenuation measuring? That attenuation would be in the CO (DSLAM)
side - not in and measured by your DSL modem.

Is your data rate measuring bit transfer rate or bit data rate?
Probably bit transfer rate - but only an educated guess -
speculation. What attenuation is being reported by status? These are
not always clear when modem does not come with detailed numerical
specs. When a telco provides a DSL modem, a modem manufacturer only
need tell your telco what those numbers measure. Some telcos would
prefer you remain ignorant.

One number that is traditionally obvious is the S/N ratio - always
above 10 dBs is necessary. Provided is enough information to make
educated guessing of what your modem might be measuring since the
manufacturer does not provide those answers in specific technical
format.

My modem report calls it Line Attenuation. I think it's also called
Loop Attenuation. It seems to mean the attenuation from inductance,
capacitance, and resistance. I suppose the modems could determine
attenuation like the cellphone commercial: "Can you hear me now?" One
sends out a 1-V signal and the other reports receiving .1 V, so they
agree it's -20 db.

I don't know the difference.

A UK ISP says you can hold a connection at 6 db. Explaining
hypothetically, they say 20 db may be required for 512k and 30 db for 2M.

My experience is that 6 dB often means a broken wire - ie. a
cracked wire held together only by its insulation. It is difficult to
explain that to phone center droids that only know from their
checklists what to do next. Tell that to a linemen, and he
immediately knows where to start his analysis.

6 dB S/N ratio (even rising as high as 8 dB) coincided with massive
noise heard on POTS phones. That example being a good DSL signal
connected to a broken telco wire. 6 dB could also indicate a low DSL
signal on a low noise line. Either way, my experience with signalling
is that 6 dB is just too low for error free communication.

We were able to achive DSL communication even at 2 dB - an analog
condition which only made it harder for those phone center droids to
understand the problem. Two and eight dB was not error free. Noise
variation was enough to create data drop outs, redundant IP packets,
and obviously slow data transfers. Once signal always exceeded 14
dB, then error free data transfers were routine.

That implies attentuation varies with frequency. Did they mean dBs
per decade? Or is that attentuation constant for all frequencies?
Again, the problem is too few facts resulting in speculation.

I've got 11 db up and 22 db down, so it seems to vary with frequency. I
don't know what modulation DSL uses, but I guess attenuation is measured
in a way to let a tech see if it could hinder communications.

Cable attenuation for UHF TV at 900 MHz or so is fierce, I recall. (I
think the flat cable had less attenuation but picked up more noise.)
I'm glad DSL has a much longer wavelength.

On past days I was getting 0 CRC errors down and 1 up. Today I'm
getting 4 down and 1 up. To a pinhead like me, that's mindboggling!

"attentuation" Always varies with Frequency....... and Distance....
sometimes it is negliable, sometimes it is significant, but that doesn't
mean it doesn't happen.....

But attentuation across a resistance is constant for all
frequencies.

Meanwhile, is it measuring attentuation for all frequencies, or
change of signal strength for all frequencies. Is the attentuation
adjusted by the modem in response to signal strength - which means
that is completely different from attenuation created by twisted pair
wires. Although atenutation on a long wire is different for various
frequencies, the additional attentuation inside the modem could be
constant for all frequencies.

But again, we don't know. Manufacturer leaves us guessing what they
mean by attentuation. We can assume it means many things - therefore
every answer is simply wild speculation.

Meanwhile, if the S/N ratio is stable at 30 dB, then I would expect
no CRC errors uploading or downloading. One problem with a 30 dB
numbers - its an average that might not see sudden and infrequent
noise. I did not realize that modem also reported packet errors.
That too (along with S/N ratios, signal strength, and latency) is a
useful tool for identifying line problems. Whatever is causing that
infrequent packet error may also cause a lower bit rate.