Low Volume on Phone Line - Load Coils
Don Kimberlin
(pc...@proctorinc.com) wrote:
> I thought that loading coils were used on extremely long >cable runs for
> equalization. On a long cable run, the capacitance >between the two wires in
> the pair gets to be enough that it is a low-reactance path for higher
> frequencies. So you can get a >subscriber pair that sounds muffled at the
> far end, because the higher the frequency of the audio, the more it is
> attenuated.
That is the psychoacoustic effect caused by the severely rolled-off
attenuation characteristic of a non-loaded pair when it has relatively
high capacitance. However, large gauge wire, let's say for example,
16 AWG (which was *very* rare in cables, but infrequently used for
some early cable carrier systems) has a surprisingly broad audio
passband. Most common experience people have is with 24 or 26 AWG and
relatively high capacitance, which was originally meant to be used
only for short distribution bits around a residential block.
Economics pressed those sizes into more and more use to the point that
they became the "norm."
> As I understand it, loading coils are hooked from each side of the line to
> ground.
Counterintuitive as it may sound, in telephone cable, "load coils"
series inductances placed at distances intended to counteract the
parallel capacitance of a set length of wire. I did a quick Web
search on "load coil" to try to provide a simple link to an
explanation, but found an amazing number of incompetent speculations
out there. It seems that if one is not on the "inside" of telephone
plant engineering, there is no limit on the amount of bovine fecal
matter idiots are willing to pile upon the topic.
"Loading" of communications wires actually predates the telephone.
Telegraphers found that the speed they could key their circuits with
was limited by a tendency for the pulses to get rounded off and
literally "smeared" from nice square waves into rounded humps of
varying voltage that obscured marks from spaces. They learned early
on that connecting "retardation coils" in series with the wire helped
to square the waves back up again. In the old intercontinental
submarine telegraph cables, a flat spiral tape was wound around the
center core to provide some "wave shaping" which was in effect
"loading." In 1898, Professor Michael Pupin did the quantitative
studies at Columbia University upon which telephone wire loading has
been based. In Europe, the practice has always had the name
"Pupinization" in honor of that work.
Over the years, a variety of "loading plans" were devised for various
sizes of wire and placements, dating back to the earliest carrier
telephone systems on aerial open-wire lines between cities. By the
1950s, cable of rather fine gauge wire with relatively high
capacitance became the economic norm for local telephone companies,
the scheme known as "H88" loading became widespread. "H" means 6000
foot (1 nautical mile) spacing and "88" means 88 millihenry
inductances.
"Loading" as used in local telephone cable is a great boon to voice
telephony, but it is a bane to broadband audio or high speed digital
transmission. However, lacking understanding, it becomes a point of
paranoia for telephone wannabes.
First off, the "6000 foot spacing" means that load coils work
improperly if they are less than 6000 feet from a termination in BOTH
directions from them. That means that unless special extra expense is
taken at the ends of the wire (by adding capacitances called "build
outs" at the ends, there must be 6000 feet on BOTH sides of a load
coil.
This means that if you are 2+ miles or less from the point where your
copper line ends, there will not be any load coils (intentionally)
placed in your pair. That's right - 2 miles or less from your CO or 2
miles or less from the end of your DLC/SLC and you won't have a
"loaded pair." That's a pretty fair distance in most local exchange
areas. (Notice the word "intentional" preceding. That's because you
*might* be in an area that was formerly rural and on a cable that was
the tail end of a once-long run now shortened and connected to a
DLC/SLC -- so a coil *could* be found on a shorter pair.
Because the "H" plan became so popular, you can see the "load coil
pots" - literally round stainless steel cans - placed every 6000 feet
along rural cable routes in the countryside. When you see one, watch
your odometer, and when you've traveled a mile, you should easily spot
the next one and so on, if traveling down a straight country road.
That plan, because it was so popular, became a very convenient one to
also locate repeaters for T-1 carrier, so the "load coil pots" today
are more often T carrier repeater housings that coil containers. Now
that you know that factoid, you know where the telco repairman is
going when he says he will "replace a T-1 repeater."
At any rate, here is one document about load coils that has a bit of
explanation about their theory and practice:
Brian Vita
> Subject: Re: Low Volume on Phone Line - Load Coils
> Date: Sat, 19 Jan 2002 21:30:52 -0500
> In article (Fri, 18 Jan 2002 09:14:27 -0800) Paul Cook
> (pc...@proctorinc.com) wrote:
>> I thought that loading coils were used on extremely long cable
>> runs for equalization. On a long cable run, the capacitance
>> between the two wires in the pair gets to be enough that it is
>> a low-reactance path for higher frequencies. So you can get a
>> subscriber pair that sounds muffled at the far end, because the
>> higher the frequency of the audio, the more it is attenuated.
> That is the psychoacoustic effect caused by the severely rolled-off
> attenuation characteristic of a non-loaded pair when it has relatively
> high capacitance. However, large gauge wire, let's say for example,
> 16 AWG (which was *very* rare in cables, but infrequently used for
> some early cable carrier systems) has a surprisingly broad audio
> passband.
Is this not a byproduct of the "skin-effect" that AC demonstrates?
That is to say, as a frequency increases, it tends to travel more
along the perimeter rather than the cross section of the cable? A
piece of 24-26AWG has a limited cross sectional area. A piece of
16AWG is much greater. I assume that these cables were also solid not
stranded.