🚀 go-pugleaf

In <1948@umriscc.isc.umr.edu> robf@mcs213k.cs.umr.edu (Rob Fugina) writes:
>Ground and neutral are exactly the same thing.  If you were to trace the

And more dangerous speculation.

It probably doesn't need to be said here, but I'll feel better
if I say it anyway:  If you don't know that you know what
you're doing, don't play around with your wiring.

There.  I do feel better.

Now, I'll lay it all out for reference.  This isn't electronics,
but maybe it will enlighten some, and help others avoid being
lit up :-)

Power wiring is always discussed in nominal RMS, peak is of
interest only when engineering insulation and protection systems.
99.999% of all wiring is done using pre-engineered components which
have their voltage and current ratings tabulated in nominal
RMS terms in catalogs and handbooks and such.

Residential service is normally provided from overhead or buried
high voltage distribution lines through a transformer serving one
or several subscribers.  The distribution is often a 3-phase Y,
but rural branches may have only one power conductor plus the
ground return.  Several other arrangements can be used to suit
the power company's needs.  Distribution lines are typically in
the 15 - 45 kV range.  The distribution circuit is protected by
overcurrent devices and ground-fault detection devices at the
substation, but it takes one heck of a fault to open those.

The subscriber transformer primary is often wired from one phase to
neutral on the distribution side, but phase-to-phase is sometimes
used.  If you have three-phase service (normally the case only
for industrial and large commercial customers) the transformer
primary is attached to all three distribution phases (and maybe neutral).
Autotransformer configurations are also used, I believe.

The residential transformer secondary is a single phase 240 volt
center tapped winding.  The center tap is coupled at the pole
to a ground rod and to the service drop neutral conductor.  This
positively limits the voltage on the service cables, relative
to ground, to a few times the nominal service voltage.  It also
ensures that a core short will draw enough current to open an
upstream protection device.

The three wire single phase service entrance cable passes through the
meter and into the customer's distribution panel.  There the neutral
is again bonded to grounds, including the frame if it is a metal
building.  This limits the voltage present on the interior wiring
relative to ground -- if the service were not gounded the whole
circuit would float.  The rules for what should and should not
be grounded, and where, and how, are complex.  Not terribly
technical, but lots of cases and exceptions have to be dealt with.
In general, neutral and ground are bonded at the service entrance
and kept separate throughout the rest of the interior wiring system.

The cable system from the transformer to the meter, together with
the transformer secondary winding, is capable of delivering some
calculable short circuit current.  There should be an overcurrent
protection device somewhere in the circuit capable of interrupting
that much current.  It is typically in the 10-20 thousand amp
range, and typical residential breaker box panels should have a main
breaker or fuse rated to handle the maximum allowble short circuit
current for a residential service drop.  If the overcurrent protection
device is a fuse there should also be a master disconnect switch.

The distribution panel supplies the two "phases" to the branch
breakers in an alternating pattern.  120V branch circuits are
assigned essentially at random to the two sides of the supply,
240V circuits use dual breakers straddling the two phases.  Due
to I^2R losses it is best to balance the loads as much as possible,
but people seldom bother.  Many breaker boxes are arranged in such
a way that only certain positions will accept a 240V circuit, or it
will only work in certain positions.  If you ever have to work inside
a breaker box you should be able to puzzle out its configuration rules
just by looking at it.

Note that the dual breakers for 240 circuits (and tripples for 3-phase)
are not just two single breakers with their handles wired together.
The trip mechanisms are connected internally so they open both
sets of contacts together more reliably than can be achieved by
tying the handles together.

Examining the interior wiring plant as a circuit, we have several
loops carying varying amounts of current.  Many of those loops involve
the neutral conductor, some are balanced 240V loads.  The neutral
current on any given 120V branch circuit is the same as the supply
current.  The overall current in the service neutral is the difference
of the currents in the two supply wires.

Interior wiring is normally designed with voltage drops in the wires
of no more than a couple of volts, so the neutral at an outlet is
normally very near the ground potential at that outlet.

Ok.  Now it get interesting.  Power system engineers earn their feed
by saving money in normal operation and saving lives and property
in fault conditions.  An electrical distribution system with a fault
looks very little like the nice little circuit model you might think
it should resemble.

Lets take three fault scenarios.

Consider an open neutral in a branch circuit.  That neutral
conductor, on the load side of the fault, now carries 120V with the load
in series limiting the current.  If that condutor is exposed
downstream of the fault a life-threatening hazard exists.  It
bears repeating that the minimum lethal dose of electric current is
in the 20 mA range.  A dangling neutral with a 100W bulb in
series can supply 50 times that (ignoring the fact that the
filament is cold).  A few volts between the cold water supply and the
drain pipe can kill a person in the bath tub.

Consider a fault to neutral -- a common scenario.  Hundreds
of amps flow through a circuit designed for 15, and you get tens
of volts dropped in the neutral.  Not only resistive voltage drop
needs to be considered -- there is a considerable inductance in
the loop.  Now if there are multiple connections between the neutral
and various "ground" points there will be currents flowing in those
grounds, which have non-negligable resistance.  different pieces
of the ground network will be at different potentials, and unless
one controls the geometry carefully (no one does), those pieces may
be close enough together for people to short themselves between them.
Yes, the overcurrent protection device will open, but it will take at
least a few milliseconds and possibly several hundred.

Consider a fault to the chassis of an appliance.  Nah, lets leave
off with the above.  In a ground fault scenario it doesn't make much
difference how the neutral is wired.  That's why we have ground fault
detectors.  By the way, in addition to the ground fault receptacles
now available (and required for bathrooms) you can get ground fault
circuit breakers for most kinds of panels.  Pricey ($40ish) but it
might come in handy sometime.  You can also get breakers with remote
control inputs or (and?) alarm outputs.  Get creative.

The important thing to remember is that you have to forget the
idea that a wire is a node.  Wires are distributed resistive+inductive
circuit elements with several potential failure modes.  And ground
is a network, not a node.  It's a lot like RF.

If any of the above comes as a surprise, you should probably limit
your electrical handy-man efforts to replacing dead parts in existing
circuits and maybe cook-booking low current (<= 20 A) branch circuits.
Find and *use* a cookbook though, don't try to wing it.  And don't
believe the salescritters at the hardware store either.  Some know
what they're talking about, but unless you know what they're talking
about you can't tell them apart from the ones who don't.

If you want to do something that is beyond your certain knowledge,
get out the phone book and call some of the small (ie, one-man)
electricians.  Chances are you can find one who will agree to check
your work (and/or plans).  You save money and still have the comfort
of a journeyman's OK.  Do I need to add that you shouldn't expect
to get his time for free?

Finally, I've probably messed up some of the above.  With probability 1,
I'm misusing some technical terms.  I'm not interested in being corrected
about trivial things, though substantive corrections are most welcome.
This is not a professional opinion.  You get what you pay for.
Be careful out there.
-- 
Steve Nuchia	      South Coast Computing Services      (713) 964-2462
	"Could we find tools that would teach their own use,
	 we should have discovered something truly beyond price."
		Socrates, in Plato's Republic