Why Negative Voltage on communication lines?
Ching
communication lines, e.g. telephone lines. Could any one give me
a brief explanation?
Thanks in advance.
K. C. Sio
Dept of Electronics Engineering
Hong Kong Polytechnic
James H. Haynes
In article <CzCzG...@hkpu01.hkp.hk>, Ching <kcsio@encserver> wrote:
>Just for curiosity, I am wondering why negative voltage is used in
>communication lines, e.g. telephone lines. Could any one give me
>a brief explanation?
I believe this has to do with electrolysis, and the fact that telephone
lines run overhead and underground and generally in places where there
is leakage to ground. I've forgotten how electrolysis works; but the
idea is that you don't want the current to carry away copper from the
wires.
belan...@wcsub.ctstateu.edu
That is essentially correct. The negative voltage with respect
to ground reverse biases the corrosion reactions.
I did not figure this out until the relevant
part of the Engineering Materials course in my BSEE curriculum.
The other way to offset corrosion is with "sacrificial" electrodes,
a technique commonly used in boats ( The "zincs" referred to in
Clancy's "Without Remorse"). In this case, the zinc corrodes
faster than the steel, protecting it, while the zinc lasts.
Hope this helps.
Gerry Belanger
Pat Trimble
K.C.,
Let me answer your question, with a question: If telephone lines are
"balanced" lines -- tip & ring, where a return lead is always furnished,
couldn't an arguement be made that positive voltage is provided, not
negative voltage?
Let me know what you think...
Pat
Gary E. Miller
: Let me answer your question, with a question: If telephone lines are
: "balanced" lines -- tip & ring, where a return lead is always furnished,
: couldn't an arguement be made that positive voltage is provided, not
: negative voltage?
Telephone lines are NOT balanced. The 'ring' is named for the ring
part of a phone plug. The 'tip' is named for the tip of the phone
plug.
The 'ring' is essentially grounded at the phone company end. The 'tip'
is -48V open circuit.
RGDS
GARY
--
---------------------------------------------------------------------------
Gary E. Miller Rellim 2680 Bayshore Pkwy, #202 Mountain View, CA 94043-1009
g...@rellim.com Tel:+1(415)964-1186 Fax:+1(415)964-1176 BBS:+1(415)964-8821
Jim....@umich.edu
Telephone lines are NOT balanced.
Not to pick nits, but I think phone lines are AC balanced, through a couple
of big inductors at the CO. They are DC unbalanced as you say.
The 'ring' is essentially grounded at the phone company end. The 'tip'
is -48V open circuit.
Isn't it the other way around?
Tony Pelliccio
> Pat Trimble (P...@ix.netcom.com) wrote:
>
> : Let me answer your question, with a question: If telephone lines are
> : "balanced" lines -- tip & ring, where a return lead is always furnished,
> : couldn't an arguement be made that positive voltage is provided, not
> : negative voltage?
>
> Telephone lines are NOT balanced. The 'ring' is named for the ring
> part of a phone plug. The 'tip' is named for the tip of the phone
> plug.
>
> The 'ring' is essentially grounded at the phone company end. The 'tip'
> is -48V open circuit.
I think they also used negative voltages to prevent corrosion of the
transmission line.
--
== Tony Pelliccio, KD1NR
== Box 1908, Prov, RI 02912 Tel. (401) 863-1880 Fax. (401) 863-2269
Patton M Turner
>I think they also used negative voltages to prevent corrosion of the
>transmission line.
Yep, that's it. Negative voltages repel negative ions, which tend to
be the ones that corrode metals. Same thing is used to protect
buildings, pipelines and underground tanks. It's called cathodic
protection, and the posative voltage is applied to a sacrifical anode.
The anode by itself is often used to protect smaller tanks.
BTW, phone lines are AC ballenced, with about 60 dB of CMRR, which is called
longitudional ballance in the industry. It can even be measured by
meters such as the sidekick. Poor CMRR causes power line harmonics
(espically 540 Hz) to be audible on the phone line.
Pat Turner FAA Telecommunications
Lorence Mlodzinski
>Just for curiosity, I am wondering why negative voltage is used in
>communication lines, e.g. telephone lines. Could any one give me
>a brief explanation?
Go back to old Alex G Bell. He used batteries for power. Carbon
transmitters varied current, recievers reproduced sound, a DC circuit.
The original crank telephones used batteries at the subscriber end.
We still refer to the voltage from the central office as exchange battery.
I doubt the original system designers thought much of balanced lines or
chemestry. They just needed a simple DC circuit. When more advanced
systems were designed they probably used Neative voltage because that's
what they used before. If you look at the history of Telcos, you'll find
that they do a lot of things because that's the way they did things before.
After 27 years in this business I've got plenty of scars to prove it.
Pat Trimble
Gary,
That "essentially grounded" ring lead is the return path to the positive
side of the central office battery -- it is NOT attached to "mother
earth," or a common ground of any kind. It is considered ground only in
reference to the tip lead which is attached to negative side of the same
central office battery. It happens to be a 48V battery, a bigger version
of the one that is in your car, based on the same Lead/Sulphuric acid
chemical reaction.
That is a balanced circuit in the sense used by the telecommunications
industry: historically the phone companies!
The tip and ring nominclature did come from the old operator's telphone
plug as you said, but that's only a naming convention used to keep the
leads straight for wiring purposes. "Fred" and "Alice" would have served
the same purpose if it had made sense to anyone.
An example of an unbalanced circuit is RS-232: multiple signal paths
with a single common return path. The reason the Bell System chose a
balanced circuit approach was to eliminate the noise induced by common
grounding arrangements: those used by the old telegraph companies that
preceded the Bell System. Telegraph circuits were thought of as
single-wire circuits. they used "mother earth" to supply the return
path. It was an economical approach that served telegraph ciruits very
well, but not the voice service sold by the telephone companies.
Technicians that use o'scopes to trouble shoot analog data circuits have
been easy prey to misunderstandings of the balanced/unbalanced concept:
they connect their scopes between either the tip or the ring lead of the
telephone company's cable pair, and "mother earth" ground. They saw lots
of "noise" on their scopes, and thought the circuit was bad. It wasn't!
What they were seeing was the difference of potential, specifically the
AC component, between the isolated telephone company cable and "mother
earth." If "mother earth" is used as a reference, the cable acts as a
big antenna, picking up all kinds of noise.
When the telephone company applys 48 Volts across a cable pair, it is
ACROSS the cable pair: tip and ring. Not from any single lead to "mother
earth" ground. That is a balanced application.
Wallace A. Ritchie
: >
: >Pat Trimble (P...@ix.netcom.com) wrote:
: >
: >: Let me answer your question, with a question: If telephone lines are
: >: "balanced" lines -- tip & ring, where a return lead is always
: furnished,
: >: couldn't an arguement be made that positive voltage is provided, not
: >: negative voltage?
: >
Both the tip and ring conductors are normally negative with respect to
earth. Previous comments regarding minimization of corrosion are exactly
the reason for this. The copper plant is arguably the LEC's most valuable
asset worth perhaps $50 - $100 billion or so. The whole idea is to
eliminate any positive voltage with respect to ground on any conductor.
: >Telephone lines are NOT balanced. The 'ring' is named for the ring
: >part of a phone plug. The 'tip' is named for the tip of the phone
: >plug.
: >
: >The 'ring' is essentially grounded at the phone company end. The 'tip'
: >is -48V open circuit.
: >
: >RGDS
: >GARY
: >--
: >-----------------------------------------------------------------------
: ----
: >Gary E. Miller Rellim 2680 Bayshore Pkwy, #202 Mountain View, CA
: 94043-1009
: >g...@rellim.com Tel:+1(415)964-1186 Fax:+1(415)964-1176
: BBS:+1(415)964-8821
: >
: Gary,
: That "essentially grounded" ring lead is the return path to the positive
: side of the central office battery -- it is NOT attached to "mother
: earth," or a common ground of any kind. It is considered ground only in
: reference to the tip lead which is attached to negative side of the same
: central office battery. It happens to be a 48V battery, a bigger version
: of the one that is in your car, based on the same Lead/Sulphuric acid
: chemical reaction.
Actually, the positive side of the central office battery plant is
normally bonded directly to earth ground. There are some floating
(i.e. not grounded) applications, however.
Both the -48V and 0V sides of the battery plant present a very low and
almost equal impedance to ground and most frequencies above DC.
The tip and ring conductors of an ordinary loop start telephone circuit
were traditionally fed through a split transformer. The tip side of the
split is traditionally fed through a 200 ohm resistor to the 0V side of
the battery (which is normally connected directly to earth). The ring
side of the split is traditionally fed through a 200 ohm resitor to the
-48V side of the battery (which also has a very low impedance to earth to
the extremely low internal resistance of the battery plant). The two
sides of the split are bypassed by a capacitor that is typically 2.16uf.
| Voice Channel |
OOOOOOOOOOOOOOOOOOOOOOOO
-------------------------- Isolation
OOOOOOO OOOO0OOOO
| |---||---| |
| > 2.16uf > |
| 200> >200 |
| ohm> >ohm |
| | | |
T 0v -48v R
_|__
/ / /
This conventional circuit presents the same AC impedance to ground for
both conductors, which is why it is called a "balanced" or "balanced to
ground" pair. Because of the near equal impedance, common mode noise i.e.
voltages that are capacitively or inductively coupled to the wire pair
will be nearly equal in both conductors and will cancel out across the
transformer. Only the differential signal from the telephone set will pass
through the transformer.
With modern equipment the tranformer, capacitor, and resistors (and
sensors not shown above) are all replaced by a SLIC (Subscriber Line
Interface Circuit) which provides the equivalent circuit through
different means.
: That is a balanced circuit in the sense used by the telecommunications
: industry: historically the phone companies!
: The tip and ring nominclature did come from the old operator's telphone
: plug as you said, but that's only a naming convention used to keep the
: leads straight for wiring purposes. "Fred" and "Alice" would have served
: the same purpose if it had made sense to anyone.
: An example of an unbalanced circuit is RS-232: multiple signal paths
: with a single common return path. The reason the Bell System chose a
: balanced circuit approach was to eliminate the noise induced by common
: grounding arrangements: those used by the old telegraph companies that
: preceded the Bell System. Telegraph circuits were thought of as
: single-wire circuits. they used "mother earth" to supply the return
: path. It was an economical approach that served telegraph ciruits very
: well, but not the voice service sold by the telephone companies.
Actually, balanced circuits are used to reduce the effects of common node
noise which are significant for thousands of feet of wire. The effects of
common returns for DC signals is a different issue. Earth is still the
return path for the ground start function on ground start trunks. The
effect of signalling through ground can be significant when there are
hundreds or thousands of circuits, each causing momentary current flows
between the battery plant and ground.
: Technicians that use o'scopes to trouble shoot analog data circuits have
: been easy prey to misunderstandings of the balanced/unbalanced concept:
: they connect their scopes between either the tip or the ring lead of the
: telephone company's cable pair, and "mother earth" ground. They saw lots
: of "noise" on their scopes, and thought the circuit was bad. It wasn't!
: What they were seeing was the difference of potential, specifically the
: AC component, between the isolated telephone company cable and "mother
: earth." If "mother earth" is used as a reference, the cable acts as a
: big antenna, picking up all kinds of noise.
: When the telephone company applys 48 Volts across a cable pair, it is
: ACROSS the cable pair: tip and ring. Not from any single lead to "mother
: earth" ground. That is a balanced application.
Actually, when you connect a single channel of a scope to one conductor of
the pair you are seeing the voltage with respect to ground of that
conductor. Because the scope has a very high input impedance, there will
be significant noise voltages. If you connect the ground clip of the
scope to the other conductor, you will normally change the impedance of
that conductor to ground. This will unbalance the pair and allow noise to
be coupled in differential mode.
If on the other hand you connect the second conductor to a second scope
channel (with equivalent impedance to ground) and operate in differential
mode (e.g. input a minus b) you will not unbalance the pair and you will
be able to see the differential mode signal.
Patton M Turner
>Both the tip and ring conductors are normally negative with respect to
>earth. Previous comments regarding minimization of corrosion are exactly
>the reason for this. The copper plant is arguably the LEC's most valuable
>asset worth perhaps $50 - $100 billion or so. The whole idea is to
>eliminate any positive voltage with respect to ground on any conductor.
The Tip is negative only because of leakage current through the phone.
Otherwise, the tip is at 0VDC and the Ring at -52 VDC. Of course, when
the phone is off hook, enough current is drawn to make tip very negative
at the customer premise, and over most of the plant. But traffic will
show most lines are used less than 5% of the time.
4 Wire circuits circuits such as program channels, voice grade data
circuits and digital circuits use sealing current to achieve the same
thing, but therie is a balanced battery on one pair and a balanced ground
on the other pair. Centertapped 600 Ohm transformers can be used to
seperate the simplex current. 20mA seems to be the standard valve for
voice band circuits. Some T1's power the repeaters this way, while
others just use it for sealing current.
Pat FAA Telecommunications
Gary E. Miller
: That "essentially grounded" ring lead is the return path to the positive
^^^^^^^^^^^
: side of the central office battery -- it is NOT attached to "mother
: reference to the tip lead which is attached to negative side of the same
: central office battery. It happens to be a 48V battery, a bigger version
: of the one that is in your car, based on the same Lead/Sulphuric acid
: chemical reaction.
: That is a balanced circuit in the sense used by the telecommunications
: industry: historically the phone companies!
Agreed, but is NOT balanced in the sense that a differential amp is balanced.
I have had many arguments with the local telco on this topic. They say
that it is OK that ring is not NEAR DC ground because it is DC balanced.
Well that DC offset saturates the input transformer on a modem and messes
up the AC CMRR (Common Mode Rejection Ratio to us non-longitudinally
balanced types). For all but a narrow passband (a few HZ to a few kHZ)
I still say that the ring is "essentially" grounded at the CO end.
: The tip and ring nominclature did come from the old operator's telphone
: plug as you said, but that's only a naming convention used to keep the
: leads straight for wiring purposes. "Fred" and "Alice" would have served
: the same purpose if it had made sense to anyone.
I will leave the semiotics to other folks....
: When the telephone company applys 48 Volts across a cable pair, it is
: ACROSS the cable pair: tip and ring. Not from any single lead to "mother
: earth" ground. That is a balanced application.
Agreed, Nor did I say so...
RGDS
GARY
--
---------------------------------------------------------------------------
Jay Ashworth
>When the telephone company applys 48 Volts across a cable pair, it is
>ACROSS the cable pair: tip and ring. Not from any single lead to "mother
>earth" ground. That is a balanced application.
Ok, fine.
Stipulated. The SLIC isolates the pair from the frame ground in the end
office. Nevertheless, the voltage in the office, relative to that frame
ground, is _-48_ volts. Again we ask the (original damned) question:
Why are telephone end offices positive ground?
David? Floyd? Anybody??
Cheers,
-- jra
--
Jay R. Ashworth High Technology Systems Consulting Ashworth
Designer Linux: The Choice of a GNU Generation & Associates
ka1fjx/4 "The difference between theory and practice +1 813 790 7592
j...@baylink.com is bigger in practice than in theory" -pds NIC: jra3
Michael Stein
Cu++ ions as well as other metal ions are positively charged.
Thus they are attracted to negatively charged objects and
repelled from positively charged objects. Do you want your
copper "repelled" off of your phone wires? (Or even the lead off
of the lead sheath?).
Bob Furtaw
wrote:
> Pat Trimble (P...@ix.netcom.com) wrote:
> ---------------------------------------------------------------------------
> Gary E. Miller Rellim 2680 Bayshore Pkwy, #202 Mountain View, CA 94043-1009
> g...@rellim.com Tel:+1(415)964-1186 Fax:+1(415)964-1176 BBS:+1(415)964-8821
Back to the original question, I believe the "negative" voltage on the
communication lines comes from a historic perspective way back from the
telegraph days when it was "discovered" that if the most negative voltage
was used on the line, less corrosion and "junk" would be attacted to this
wire. This was in the days when the ground line actually returned through
the physical ground. Through the years in order to maintain compatability,
the issue remained. It's even made it through to RS-232 standards.
Of course balanced lines and other techniques improved the situation, but
much of the terminology and references die hard. Just MHO based on
historical knowledge.
Bob Furtaw - W8IL fur...@comm.mot.com
CFI-A/G/I/MEI, CGI-A/I
All disclaimers ever written by anybody apply. :-)
adams,john
Gary E. Miller <g...@aplatform.com> wrote:
>Pat Trimble (P...@ix.netcom.com) wrote:
>
>: Let me answer your question, with a question: If telephone lines are
>: "balanced" lines -- tip & ring, where a return lead is always furnished,
>: couldn't an arguement be made that positive voltage is provided, not
>: negative voltage?
>
>Telephone lines are NOT balanced. The 'ring' is named for the ring
>part of a phone plug. The 'tip' is named for the tip of the phone
>plug.
>
>The 'ring' is essentially grounded at the phone company end. The 'tip'
>is -48V open circuit.
Not exactly, Ring is at negative potential while Tip is at or close to ground.
This allowed old corded switchboards to test busy on a line by touching the
"Sleeve" lead of the line appearance
with the Tip momentarily and listening on the cord circuit
headset for the distinctive click.
--
John (Jack) Adams | Bellcore NVC 2Z-220
(908) 758-5372 {Voice} | (908) 758-4389 {Facsimile}
jad...@cc.bellcore.com | (914) 644-2831 {NYNEX S&T}
-- The weather is here, I wish you were beautiful -- Jimmy Buffett
Wallace A. Ritchie
: writ...@gate.net (Wallace A. Ritchie) writes:
: >Both the tip and ring conductors are normally negative with respect to
: >earth. Previous comments regarding minimization of corrosion are exactly
: >the reason for this. The copper plant is arguably the LEC's most valuable
: >asset worth perhaps $50 - $100 billion or so. The whole idea is to
: >eliminate any positive voltage with respect to ground on any conductor.
: The Tip is negative only because of leakage current through the phone.
: Otherwise, the tip is at 0VDC and the Ring at -52 VDC. Of course, when
: the phone is off hook, enough current is drawn to make tip very negative
: at the customer premise, and over most of the plant. But traffic will
: show most lines are used less than 5% of the time.
Or leakage in the tip conductor, protectors at the CO, protectors at the
entrance, etc. The point is that positive voltages are not present to
prevent corrosion to the cable. Actually, the open circuit voltages
across tip and ring can range from 6.5 Volts (older NCTE) to 105 Volts
(channels with loop extenders) although 48 Volts is most common.
: 4 Wire circuits circuits such as program channels, voice grade data
: circuits and digital circuits use sealing current to achieve the same
: thing, but therie is a balanced battery on one pair and a balanced ground
: on the other pair. Centertapped 600 Ohm transformers can be used to
: seperate the simplex current. 20mA seems to be the standard valve for
: voice band circuits. Some T1's power the repeaters this way, while
: others just use it for sealing current.
Sealing currents are a different matter. The purpose of sealing current
is to maintain the quality of the metallic interconnections, particularly
where wire wrapped or soldered connections are not used. A
continual current-flow helps maintain the integrity of the
connections (Otherwise you would just put -48 on both conductors).
Sealing currents can also confirm the metallic integrity of the connection.
For T1 span's, the voltages used in simplex feeds is often well above 100
volts and may include both positive and negative (with respect to ground)
batteries. As I recall, the low power repeaters use a constant current
feed of 50ma at whatever voltage is required to maintain it. Each one way
repeater draws its power in a fashion that drops a constant voltage. The
repeaters are traditionally spaced every 6000 feet (not coincidently the
same spacing as loading coils).
W. A. Ritchie
Ft. Lauderdale, Florida
Wallace A. Ritchie
: Pat Trimble (P...@ix.netcom.com) wrote:
: : That "essentially grounded" ring lead is the return path to the positive
: ^^^^^^^^^^^
: : side of the central office battery -- it is NOT attached to "mother
: : earth," or a common ground of any kind. It is considered ground only in
: : reference to the tip lead which is attached to negative side of the same
: : central office battery. It happens to be a 48V battery, a bigger version
: : of the one that is in your car, based on the same Lead/Sulphuric acid
: : chemical reaction.
: : That is a balanced circuit in the sense used by the telecommunications
: : industry: historically the phone companies!
: Agreed, but is NOT balanced in the sense that a differential amp is balanced.
: I have had many arguments with the local telco on this topic. They say
: that it is OK that ring is not NEAR DC ground because it is DC balanced.
You may agree but this is wrong. The positive side of the battery plant
may be earth grounded (usually) or floating. Also the circuit is balanced in
the same sense that a differential amp is balanced.
: Well that DC offset saturates the input transformer on a modem and messes
: up the AC CMRR (Common Mode Rejection Ratio to us non-longitudinally
: balanced types). For all but a narrow passband (a few HZ to a few kHZ)
: I still say that the ring is "essentially" grounded at the CO end.
It is not the DC offset that saturates the transformer but the rather the
DC current flow through the transformer which is normally between 23 ma and
85 ma. Preventing saturation generally requires a relatively large
device. On the other hand, some designs use a "current eater" or virtual
inductor with a DC blocking capacitor which solves this problem and
allows the use a small transformer.
The DC offset to ground has nothing to do with the CMRR across the
transformer. What affects the CMRR is any imbalance to ground on either
the tip side or the ring side of the circuit.
<deleted>
: : When the telephone company applys 48 Volts across a cable pair, it is
: : ACROSS the cable pair: tip and ring. Not from any single lead to "mother
: : earth" ground. That is a balanced application.
: Agreed, Nor did I say so...
: RGDS
: GARY
You may agree but this is also wrong! It is the equal impedance to ground
of each conductor that makes the pair balanced. You can apply -48V of a
(positive ground) battery to one lead through a 2K resistor and 0V (or
ground) to the other lead and you will have an unbalanced pair with an
open circuit voltage of 48V.
Wallace A. Ritchie
: P...@ix.netcom.com (Pat Trimble) writes:
: >When the telephone company applys 48 Volts across a cable pair, it is
: >ACROSS the cable pair: tip and ring. Not from any single lead to "mother
: >earth" ground. That is a balanced application.
: Ok, fine.
: Stipulated. The SLIC isolates the pair from the frame ground in the end
: office. Nevertheless, the voltage in the office, relative to that frame
: ground, is _-48_ volts. Again we ask the (original damned) question:
Not OK. The SLIC may be isolated or not. If the office battery is
isolated, both conductors will be floating with respect to ground. There
is no issue. Call tip ring and visa versa.
: Why are telephone end offices positive ground?
: David? Floyd? Anybody??
When one side of the battery is earth grounded it will always be the
positive side to prevent metallic corrosion as pointed out by many.
Wallace A. Ritchie
: Back to the original question, I believe the "negative" voltage on the
: communication lines comes from a historic perspective way back from the
: telegraph days when it was "discovered" that if the most negative voltage
: was used on the line, less corrosion and "junk" would be attacted to this
: wire. This was in the days when the ground line actually returned through
: the physical ground.
Even when a separate return line is used, any voltage above ground on
either conductor creates the potential (NPI) for metallic corrosion.
: Through the years in order to maintain compatability,
: the issue remained. It's even made it through to RS-232 standards.
^^^^^^^^^^^^^^^
What do you mean? RS-232 has both positive and negative signals around a
common signal ground which may be isolated from earth.
Floyd Davidson
>Pat Trimble (P...@ix.netcom.com) wrote:
>: That "essentially grounded" ring lead is the return path to the positive
> ^^^^^^^^^^^
>: side of the central office battery -- it is NOT attached to "mother
>: earth," or a common ground of any kind. It is considered ground only in
>: reference to the tip lead which is attached to negative side of the same
>: central office battery. It happens to be a 48V battery, a bigger version
>: of the one that is in your car, based on the same Lead/Sulphuric acid
>: chemical reaction.
>
>: That is a balanced circuit in the sense used by the telecommunications
>: industry: historically the phone companies!
>Agreed, but is NOT balanced in the sense that a differential amp is balanced.
>I have had many arguments with the local telco on this topic. They say
>that it is OK that ring is not NEAR DC ground because it is DC balanced.
Somebody is feeding you a line... it is not DC balanced, it is AC balanced.
>Well that DC offset saturates the input transformer on a modem and messes
The input transformer on a modem is designed to handle the line current,
it does not saturate.
>up the AC CMRR (Common Mode Rejection Ratio to us non-longitudinally
If it does your modem is exceedingly ill designed.
>balanced types). For all but a narrow passband (a few HZ to a few kHZ)
>I still say that the ring is "essentially" grounded at the CO end.
The ring is at DC ground potential in the Central Office. It is not
likely to be exactly at ground potential anywhere else, just close.
A twisted pair telephone cable is a balanced transmission line for
AC, and the bandwidth is not hardly restricted to a narrow passband
of a few KHz, but rather to several MegaHertz.
>: When the telephone company applys 48 Volts across a cable pair, it is
>: ACROSS the cable pair: tip and ring. Not from any single lead to "mother
>: earth" ground. That is a balanced application.
>Agreed, Nor did I say so...
It is in fact 52Vdc to ground.
Floyd
--
fl...@hayes.alaska.edu A guest on the Institute of Marine Science computer
Salcha, Alaska system at the University of Alaska at Fairbanks.
Bob Furtaw
Ritchie) wrote:
> Bob Furtaw (fur...@comm.mot.com) wrote:
>
> : Back to the original question, I believe the "negative" voltage on the
> : communication lines comes from a historic perspective way back from the
> : telegraph days when it was "discovered" that if the most negative voltage
> : was used on the line, less corrosion and "junk" would be attacted to this
> : wire. This was in the days when the ground line actually returned through
> : the physical ground.
>
> Even when a separate return line is used, any voltage above ground on
> either conductor creates the potential (NPI) for metallic corrosion.
Agreed
> : Through the years in order to maintain compatability,
> : the issue remained. It's even made it through to RS-232 standards.
> ^^^^^^^^^^^^^^^
> What do you mean? RS-232 has both positive and negative signals around a
> common signal ground which may be isolated from earth.
I was only referrring to the terminology where "logic 1" is the most
negative voltage level as specified in EIA RS-232. ...Which can be traced
back to "loop" terminology...which can be traced to original telegraph
wiring.
> W. A. Ritchie
> Ft. Lauderdale, Florida
Bob Furtaw - W8IL fur...@comm.mot.com
Carl Knoblock
: In <CzCzG...@hkpu01.hkp.hk> kcsio@encserver (Ching) writes:
:
: >
: >Just for curiosity, I am wondering why negative voltage is used in
: >communication lines, e.g. telephone lines. Could any one give me
: >a brief explanation?
: Let me answer your question, with a question: If telephone lines are
: couldn't an arguement be made that positive voltage is provided, not
: negative voltage?
:
: Let me know what you think...
:
: Pat
Interesting hypotesis, but the fact is that the positive side of the
battery in the central office IS grounded. Not all lines are balanced in
the idle state, either. Lines connecting to PBXs are usually "open tip"
with negative battery on the ring, only. They change to balanced when a
connection is set up, even to dial tone.
--
Carl Knoblock - Metro Apple Computer Hobbyists (MACH)
"The Dead Computer Society" (Apple IIgs)
ckn...@gonix.com | c.kno...@genie.geis.com
Wallace A. Ritchie
: In article <3aq5e4$3...@aplatform.com> g...@aplatform.com (Gary E. Miller)
: writes:
: >Pat Trimble (P...@ix.netcom.com) wrote:
: >: That "essentially grounded" ring lead is the return path to the positive
: > ^^^^^^^^^^^
: >: That is a balanced circuit in the sense used by the telecommunications
: >Agreed, but is NOT balanced in the sense that a differential amp is
: > balanced.
: >I have had many arguments with the local telco on this topic. They say
: >that it is OK that ring is not NEAR DC ground because it is DC balanced.
: Somebody is feeding you a line... it is not DC balanced, it is AC balanced.
Wow! What connectivity on the internet. In addition to 84 countries we
now seem to have connections to other universes that don't obey the laws
of physics we have here in my universe.
On planet Earth (named after earth ground) in my universe, longitudinal
balance is the relative impedance to ground of the two conductors of a
tranmission line and the terminals that attach to it at any point. This
impedance normally includes both real (resistive) and imaginary
(reactive) components.
In my universe we use balanced transmission pair, and balanced terminals
so that longitudally coupled noise currents (i.e. noise coupled to both
conductors) do not appear as a metalic voltage (i.e. across tip and ring
at the terminal).
We have objective standards for longitudinal balance for CPE and the
LEC's maintain there own objective standards for their outside plant and
network terminals. In some cases CPE will be at both terminals with the
LEC providing only the outside plant (for example an OPX line from a PBX).
Regulatory balance requirements (FCC Part 68.310) for CPE in the US apply
only to the frequency ranges of 200hz to 4000hz for analog voice and
200hz to 1.544mhz for digital. Other standards extend the lower boundary
down to 60hz.
Balance itself refers to impedance. Measuring balance requires a
specificiation of the frequency. At DC, the reactive components are zero
and only the resistive component of the impedance affects the total
impedance. At AC both resistive and reactive (capacitive and inductive)
components of the impedance are involved. In the real world the resistive
and inductive components are normally separately balanced. It is possible,
however, to have a balanced (to some objective standard) impedance over
some arbitrary frequency range when they are not. In this case there must
be some reactive component providing the balancing effect to offset the
unbalanced resistivie component.
As a practical matter, telephone terminals (including the CO) have the
same resistive impedance to ground on both tip and ring, in other words
they are DC balanced. When the battery is floating the resistance to
ground will be very high reflecting only leakage to ground. When one side
(usually positive) of the battery is grounded, the internal resistance of
the battery plant to ground from that side is effectively zero. The
resistive component of the impedance to ground from the other side will be
the internal resistance of the battery (typically < 0.01 ohm). The LEC (or
the PBX manufacturer) will want to limit the current flow (AC or DC) in
both tip and ring so there will normally be equal resistance in both the
tip side and the ring side. The resistive component of the impedance to
ground from a grounded battery feed will normally be about 200 ohms on
each side. In fact, resistors used in real world battery feeds are matched
to 0.5 % or better.
If the battery is ungrounded, then the resistive component of the
impedance to ground will be very high on both sides. The relative
resistance in tip or ring will not much affect the balance because it
will be very small relative to the resistance to ground through leakage
paths.
BTW the metallic current flow in tip and ring due to the battery is signal.
The signal by definition is unbalanced or it won't appear across tip and
ring. The fact that there is a DC component of the signal has absolutely
nothing to do with the longitudal balance. Neither does the fact that the
DC current in one conductor is different from the DC current in the other
conductor. What matters it the metallic voltage across tip and ring and
the relative impedance to ground from tip and ring.
You can measure DC balance with a resitive networks as follows:
Ring Tip
< <
Rr < < Rt
< <
|_______|
|
<
< Rg
<
|__
___ |- s1
_|_
/ / /
Rr and Rt should be matched to 0.5% for a resonable measure. (Otherwise
this network will be unbalanced when connected to ground). Rg can be about
450 ohms. If you connect this network across tip and ring, the metallic
voltage across tip and ring will be x where x is a function of the office
battery and total loop resistance. The line will register off-hook and you
will receive dialtone. If you momentarily press s1 there will be no change
in the metallic voltage across Tip and Ring for a line that is resistively
balanced to ground. There will be a change in the current in the TIP and
the current in RING unless the battery if floating. The voltage across Rt
and Rr will also be different unless the battery is floating (or both
positive and negative batteries with common ground are used (very rare).
[Note that this network may interfere with the signaling because the CO
normally expects a DC current through ring that is equal and opposite of
that through tip. Grounding the balance point causes the feed current to
return through ground where it may not be measured by a solid state SLIC].
: <more deleted>
: It is in fact 52Vdc to ground.
: Floyd
: --
: fl...@hayes.alaska.edu A guest on the Institute of Marine Science computer
: Salcha, Alaska system at the University of Alaska at Fairbanks.
The industry uses the term 48V to refer a nominal battery voltage of 48
volts plus or minus 4 volts, i.e. 44 - 52 Volts. 52 Volts normally
represents the fully charged battery plant state within the nominal
range.
Jay Ashworth
}The ring is at DC ground potential in the Central Office. It is not
}likely to be exactly at ground potential anywhere else, just close.
}A twisted pair telephone cable is a balanced transmission line for
}AC, and the bandwidth is not hardly restricted to a narrow passband
}of a few KHz, but rather to several MegaHertz.
Modulo a loading coil or two; see ADSL. :-)
Floyd Davidson
>Floyd Davidson (fl...@hayes.ims.alaska.edu) wrote:
>: In article <3aq5e4$3...@aplatform.com> g...@aplatform.com (Gary E. Miller)
>: writes:
>: >Pat Trimble (P...@ix.netcom.com) wrote:
>: >: That "essentially grounded" ring lead is the return path to the positive
>: > ^^^^^^^^^^^
>: > <stuff deleted>
>: >: That is a balanced circuit in the sense used by the telecommunications
>: >: industry: historically the phone companies!
>: >Agreed, but is NOT balanced in the sense that a differential amp is
>: > balanced.
>: >I have had many arguments with the local telco on this topic. They say
>: >that it is OK that ring is not NEAR DC ground because it is DC balanced.
>
>: Somebody is feeding you a line... it is not DC balanced, it is AC balanced.
>
>Wow! What connectivity on the internet. In addition to 84 countries we
>now seem to have connections to other universes that don't obey the laws
>of physics we have here in my universe.
Chuckle. When that appears to be the case, usually it means you
don't understand either the law or the application...
>In my universe we use balanced transmission pair, and balanced terminals
>so that longitudally coupled noise currents (i.e. noise coupled to both
>conductors) do not appear as a metalic voltage (i.e. across tip and ring
>at the terminal).
Yep, that is how AC signals are sent down the twisted pair balanced
transmission line.
Nope, that is NOT how it is used for DC by the telco. Consider
the current loop on/off hook supervision. Then consider Ground
Start trunks, then think about EMX or DX signaling (or any other
kind of composite signaling). Hmmm... how about the voltage used
to power various facilities such as T1 repeaters.
I don't think the pair is considered a DC balanced device by the
telco.
You applied all the right ideas to the wrong places mostly because
of terminology variations. Certainly the cable itself is required
to have equal (AC) resistive components for the tip and ring, but
that is not the DC balance being asked about (at least _I_ don't
think that was the question...). The telco uses the cable for a
variety of DC applications that are NOT balanced. (And hence my
original reply to the statement that the DC voltage on the ring
can be non zero because the line is DC balanced. The DC can be non
zero because it is AC balanced and not DC balanced.)
>Regulatory balance requirements (FCC Part 68.310) for CPE in the US apply
>only to the frequency ranges of 200hz to 4000hz for analog voice and
>200hz to 1.544mhz for digital. Other standards extend the lower boundary
>down to 60hz.
>
>Balance itself refers to impedance. Measuring balance requires a
>specificiation of the frequency.
That should have told you something...
> At DC, the reactive components are zero
>and only the resistive component of the impedance affects the total
>impedance.
At DC there is no such thing as impedance. And the AC resistive component
of the impedance may not be the same as the DC resistance. Consider for
a moment the standard telco quiet term. A 2.15 mfd cap in series with
a 600 ohm resister. AC resistive component equals 600 ohms while the
DC resistive component is the leakage resistance of the capacitor...
>At AC both resistive and reactive (capacitive and inductive)
>components of the impedance are involved. In the real world the resistive
>and inductive components are normally separately balanced. It is possible,
>however, to have a balanced (to some objective standard) impedance over
>some arbitrary frequency range when they are not. In this case there must
>be some reactive component providing the balancing effect to offset the
>unbalanced resistivie component.
Using reactive components to balance a resistive component is not
a broadband solution.
>As a practical matter, telephone terminals (including the CO) have the
>same resistive impedance to ground on both tip and ring, in other words
>they are DC balanced. When the battery is floating the resistance to
In any given case it might be balanced, but it is NOT a
requirement for proper operation of a local loop. The AC component
(vf signals) are isolated from the DC component by a repeat coil,
and what is done DC-wise on the other side of that repeat coil
(a split winding transformer) has no effect on the operation of
the twisted pair cable as a balanced transmission line. It may
in fact be that under normal circumstances the resistance to
ground on both sides is equal... but it also may be adjusted
in some cases for short, medium, or long loops, and that is
not necessarily done equally to both sides because all that
matters is total loop resistance.
>BTW the metallic current flow in tip and ring due to the battery is signal.
>The signal by definition is unbalanced or it won't appear across tip and
>ring. The fact that there is a DC component of the signal has absolutely
>nothing to do with the longitudal balance. Neither does the fact that the
>DC current in one conductor is different from the DC current in the other
>conductor. What matters it the metallic voltage across tip and ring and
>the relative impedance to ground from tip and ring.
The above is correct. The telco usage of DC on the cable has
nothing to do with longitudal balance. It is used often in an
unbalanced manner with no effect on the impedance or balance of
the line.
>: <more deleted>
>
>: It is in fact 52Vdc to ground.
...
>The industry uses the term 48V to refer a nominal battery voltage of 48
>volts plus or minus 4 volts, i.e. 44 - 52 Volts. 52 Volts normally
>represents the fully charged battery plant state within the nominal
>range.
A fully charged battery plant state is 48 VDC. 52 VDC is the voltage
when being float charged. Non-telco people rarely ever see anything
other than 52VDC, but it happens that even higher is possible when
a battery plant is put on "equalize" charge.
CHARLES ALLEN
corrosion problems. This is similar to the cathodic protection systems
used on ships.
Wallace A. Ritchie
: In article <3b2s4h$s...@tequesta.gate.net> writ...@gate.net (Wallace A. Ritchie) writes:
: >Floyd Davidson (fl...@hayes.ims.alaska.edu) wrote:
: >: In article <3aq5e4$3...@aplatform.com> g...@aplatform.com (Gary E. Miller)
: >: writes:
: >: >Pat Trimble (P...@ix.netcom.com) wrote:
: >: >: That "essentially grounded" ring lead is the return path to the positive
: >: > ^^^^^^^^^^^
: >: > <stuff deleted>
: >: >: That is a balanced circuit in the sense used by the telecommunications
: >: >: industry: historically the phone companies!
: >: >Agreed, but is NOT balanced in the sense that a differential amp is
: >: > balanced.
: >: >I have had many arguments with the local telco on this topic. They say
: >: >that it is OK that ring is not NEAR DC ground because it is DC balanced.
: >
: >: Somebody is feeding you a line... it is not DC balanced, it is AC balanced.
: >
: >Wow! What connectivity on the internet. In addition to 84 countries we
: >now seem to have connections to other universes that don't obey the laws
: >of physics we have here in my universe.
: Chuckle. When that appears to be the case, usually it means you
: don't understand either the law or the application...
Yes Floyd, one or both of us is a little confused here. Or more likely we
have a TERMination problem.
: >In my universe we use balanced transmission pair, and balanced terminals
: >so that longitudally coupled noise currents (i.e. noise coupled to both
: >conductors) do not appear as a metalic voltage (i.e. across tip and ring
: >at the terminal).
: Yep, that is how AC signals are sent down the twisted pair balanced
: transmission line.
: Nope, that is NOT how it is used for DC by the telco. Consider
: the current loop on/off hook supervision. Then consider Ground
: Start trunks, then think about EMX or DX signaling (or any other
: kind of composite signaling). Hmmm... how about the voltage used
: to power various facilities such as T1 repeaters.
: I don't think the pair is considered a DC balanced device by the
: telco.
For loop start, tip and ring are balanced all the way to DC in both
states. Same for loop reverse battery (DID). For ground start, the pair
is not balanced in the idle condition as each wire is, in effect, a
separate circuit fed from different batteries, (or different impedance
from the same battery when the CO battery is looped by the PBX. When the
ground start trunk converts to loop operation with a voice path it becomes
balanced down to DC. Hi-Lo and Wet Dry both balanced down to DC. CX and
DX signalling. Hmmmm....... Got us both here Floyd. (See Below) Anybody
else still with us hang on!
: You applied all the right ideas to the wrong places mostly because
: of terminology variations. Certainly the cable itself is required
: to have equal (AC) resistive components for the tip and ring, but
: that is not the DC balance being asked about (at least _I_ don't
: think that was the question...). The telco uses the cable for a
^^^^^^^^^^^^^^
: variety of DC applications that are NOT balanced. (And hence my
: original reply to the statement that the DC voltage on the ring
: can be non zero because the line is DC balanced. The DC can be non
: zero because it is AC balanced and not DC balanced.)
The original question was regarding loop start pairs and why the TELCO's
supposed statements that they are DC balanced was wrong. I still maintain
that they are typically balanced down to DC. The resistive component of
the impedance to ground in tip and ring are equal. We'll get to CX and DX
trunks in a minute.
: >Regulatory balance requirements (FCC Part 68.310) for CPE in the US apply
: >only to the frequency ranges of 200hz to 4000hz for analog voice and
: >200hz to 1.544mhz for digital. Other standards extend the lower boundary
: >down to 60hz.
: >
: >Balance itself refers to impedance. Measuring balance requires a
: >specificiation of the frequency.
: That should have told you something...
It tells me that the balance to ground at DC is irrelevant to Part 68. I
only have worry about 200 hz and up. (60hz if I'm designing to EIA
or T.1 standards.) This means that if the CPE is grounded (for example it
supplies battery and ground for OPX or Loop Reverse Battery (DID)), I can
be as unbalanced at DC as I want provided that I restore balance at 200hz
(or 60hz) and above.
: > At DC, the reactive components are zero
: >and only the resistive component of the impedance affects the total
: >impedance.
: At DC there is no such thing as impedance. And the AC resistive component
: of the impedance may not be the same as the DC resistance. Consider for
: a moment the standard telco quiet term. A 2.15 mfd cap in series with
: a 600 ohm resister. AC resistive component equals 600 ohms while the
: DC resistive component is the leakage resistance of the capacitor...
I think of impedance at DC as being equal to resistance, impedance being
a generalization that includes both resistive and reactive (real and
imaginary) components. Saying that at 0 hz impedance disappears and is
replaced by resistance is ok with me so the above is correct. In the real
world the imaginary components matter because we have noise at AC.
: >At AC both resistive and reactive (capacitive and inductive)
: >components of the impedance are involved. In the real world the resistive
: >and inductive components are normally separately balanced. It is possible,
: >however, to have a balanced (to some objective standard) impedance over
: >some arbitrary frequency range when they are not. In this case there must
: >be some reactive component providing the balancing effect to offset the
: >unbalanced resistivie component.
: Using reactive components to balance a resistive component is not
: a broadband solution.
Yup. But I think this is the case with CX and DX trunks. I believe that
the DC resistance to ground in this case is not balanced. AC balance is
maintained by feeding battery or ground through a relatively high
inductance on both sides and adding a compensating network. In the case of
trunks you can afford to have manual adjustments like this.
: >As a practical matter, telephone terminals (including the CO) have the
: >same resistive impedance to ground on both tip and ring, in other words
: >they are DC balanced. When the battery is floating the resistance to
: In any given case it might be balanced, but it is NOT a
: requirement for proper operation of a local loop. The AC component
: (vf signals) are isolated from the DC component by a repeat coil,
: and what is done DC-wise on the other side of that repeat coil
: (a split winding transformer) has no effect on the operation of
: the twisted pair cable as a balanced transmission line. It may
: in fact be that under normal circumstances the resistance to
: ground on both sides is equal... but it also may be adjusted
: in some cases for short, medium, or long loops, and that is
: not necessarily done equally to both sides because all that
: matters is total loop resistance.
1. The impedance from T to ground and R to ground must be balanced (i.e.
equal) over the frequency of interest. I think we completely agree on
that.
2. If we feed ground to T and battery to R through equal resistances we
will be balanced to ground at DC. If we feed through unequal resistances
to ground (for example connecting Tip directly to ground) we will be
unbalanced at DC. I am making no distinction here between AC and DC
resistance because there can be no blocking capacitor in the DC feed.
3. Being unbalanced to ground at DC is OK as long as we are balanced in
the off-hook state in the frequencies of interest. If we are unbalanced
at DC (i.e. resistively) we must have some compensating unbalance with
reactive components.
4. In the case of repeating coils, these are normally balanced, i.e. the
inductance of each winding is equal. If the inductance of the coils is
high it will lessen the effect of minor imbalances in the DC resistances
by raising the impedance to ground on both sides and thus minimizing ther
differences. But the inductances of repeating coils and transformers is
not nearly large enough to cancel at low VF the effects of unbalanced DC
resistances.
: >BTW the metallic current flow in tip and ring due to the battery is signal.
: >The signal by definition is unbalanced or it won't appear across tip and
: >ring. The fact that there is a DC component of the signal has absolutely
: >nothing to do with the longitudal balance. Neither does the fact that the
: >DC current in one conductor is different from the DC current in the other
: >conductor. What matters it the metallic voltage across tip and ring and
: >the relative impedance to ground from tip and ring.
: The above is correct. The telco usage of DC on the cable has
: nothing to do with longitudal balance. It is used often in an
: unbalanced manner with no effect on the impedance or balance of
: the line.
I think this is where we have the terminology problem. I am interpreting
balance to mean longitudinal balance i.e. the relative impedance to
ground of Tip and Ring. There is another use of balance to refer to
current flow.
BOTH DC AND AC CURRENTS ARE UNBALANCED. In a normal loop the DC flows in
one direction in Tip and the opposite direction to Ring. Even when the
current has the same magnitude it flows in opposite directions. The same
is true for the AC currents which are also unbalanced. When positive flows
down tip negative flows back on ring. The common mode noise currents, in
contrast, flow in the same direction and have the same magnitude in a
balanced circuit.
The DC current is not balanced because it has the same magnitude and
opposite sign. It is balanced in this sense only if the magnitude and
sign are both the same. This was my point about calling the DC battery
feed a signal.
In terms of current flow, the AC signal will be an unbalanced flow and the
talk battery will be an unblanced flow. The current through tip will
normally (in loop start) be opposite in sign to the current through ring.
In some arrangements there may be no DC current in either tip or ring, DC
current in tip but not ring, DC current in ring but not tip, or any
arbitrary DC current in either.
: >: <more deleted>
: >
: >: It is in fact 52Vdc to ground.
: ...
: >The industry uses the term 48V to refer a nominal battery voltage of 48
: >volts plus or minus 4 volts, i.e. 44 - 52 Volts. 52 Volts normally
: >represents the fully charged battery plant state within the nominal
: >range.
: A fully charged battery plant state is 48 VDC. 52 VDC is the voltage
: when being float charged. Non-telco people rarely ever see anything
: other than 52VDC, but it happens that even higher is possible when
: a battery plant is put on "equalize" charge.
: Floyd
In the case of a PBX providing an OPX the maximum allowed voltage per FCC
68.306 is -56.5 Volts. Also, at least for non-telcos FCC part 68 requires
that the voltages on tip and ring be negative with respect to ground.
: --
: fl...@hayes.alaska.edu A guest on the Institute of Marine Science computer
: Salcha, Alaska system at the University of Alaska at Fairbanks.
Thanks for the workout Floyd!
Wally Ritchie
Ft. Lauderdale, Florida
Floyd Davidson
can do without a lot of the baggage, so I'm chopping off a lot of the
original discussion and attributions, and it is still far too long.
(Which probably says that Wally and I are a couple windbags...)
>: >: >I have had many arguments with the local telco on this topic. They say
>: >: >that it is OK that ring is not NEAR DC ground because it is DC balanced.
>: >
>: >: Somebody is feeding you a line... it is not DC balanced, it is AC balanced.
writ...@gate.net (Wallace A. Ritchie) writes:
>The original question was regarding loop start pairs and why the TELCO's
>supposed statements that they are DC balanced was wrong. I still maintain
>that they are typically balanced down to DC. The resistive component of
>the impedance to ground in tip and ring are equal. We'll get to CX and DX
>trunks in a minute.
And I maintain that the DC circuit on a cable pair is not a balanced
device. It is isolated from the AC circuit, which is using the cable
as a balanced device.
>only have worry about 200 hz and up. (60hz if I'm designing to EIA
>or T.1 standards.) This means that if the CPE is grounded (for example it
>supplies battery and ground for OPX or Loop Reverse Battery (DID)), I can
>be as unbalanced at DC as I want provided that I restore balance at 200hz
>(or 60hz) and above.
And on occasion, Wally agrees with me...
>: >At AC both resistive and reactive (capacitive and inductive)
>: >components of the impedance are involved. In the real world the resistive
>: >and inductive components are normally separately balanced. It is possible,
>: >however, to have a balanced (to some objective standard) impedance over
>: >some arbitrary frequency range when they are not. In this case there must
>: >be some reactive component providing the balancing effect to offset the
>: >unbalanced resistivie component.
[>Floyd Davidson (fl...@sanford.ims.alaska.edu) wrote:]
>: Using reactive components to balance a resistive component is not
>: a broadband solution.
>
>Yup. But I think this is the case with CX and DX trunks. I believe that
>the DC resistance to ground in this case is not balanced. AC balance is
>maintained by feeding battery or ground through a relatively high
>inductance on both sides and adding a compensating network. In the case of
>trunks you can afford to have manual adjustments like this.
DC signaling is applied to a circuit in exactly the same manner as
loop current on a dial line. It does not involve feeding battery/ground
through high inductances on both sides or adding comp networks. It is
done with a split winding repeat coil, just like on a dial line.
>: >As a practical matter, telephone terminals (including the CO) have the
>: >same resistive impedance to ground on both tip and ring, in other words
>: >they are DC balanced. When the battery is floating the resistance to
[>Floyd Davidson (fl...@sanford.ims.alaska.edu) wrote:]
>: In any given case it might be balanced, but it is NOT a
>: requirement for proper operation of a local loop. The AC component
>: (vf signals) are isolated from the DC component by a repeat coil,
>: and what is done DC-wise on the other side of that repeat coil
>: (a split winding transformer) has no effect on the operation of
>: the twisted pair cable as a balanced transmission line. It may
>: in fact be that under normal circumstances the resistance to
>: ground on both sides is equal... but it also may be adjusted
>: in some cases for short, medium, or long loops, and that is
>: not necessarily done equally to both sides because all that
>: matters is total loop resistance.
>
>1. The impedance from T to ground and R to ground must be balanced (i.e.
>equal) over the frequency of interest. I think we completely agree on
>that.
Yes.
>2. If we feed ground to T and battery to R through equal resistances we
>will be balanced to ground at DC. If we feed through unequal resistances
>to ground (for example connecting Tip directly to ground) we will be
>unbalanced at DC. I am making no distinction here between AC and DC
>resistance because there can be no blocking capacitor in the DC feed.
Yes.
>3. Being unbalanced to ground at DC is OK as long as we are balanced in
>the off-hook state in the frequencies of interest. If we are unbalanced
>at DC (i.e. resistively) we must have some compensating unbalance with
>reactive components.
No.
>4. In the case of repeating coils, these are normally balanced, i.e. the
>inductance of each winding is equal. If the inductance of the coils is
>high it will lessen the effect of minor imbalances in the DC resistances
>by raising the impedance to ground on both sides and thus minimizing ther
>differences. But the inductances of repeating coils and transformers is
>not nearly large enough to cancel at low VF the effects of unbalanced DC
>resistances.
The inductance of the repeat coils (which ARE transformers) is not
relevant to DC imbalance.
>: >ring. The fact that there is a DC component of the signal has absolutely
>: >nothing to do with the longitudal balance. Neither does the fact that the
>: >DC current in one conductor is different from the DC current in the other
>: >conductor. What matters it the metallic voltage across tip and ring and
>: >the relative impedance to ground from tip and ring.
>
[>Floyd Davidson (fl...@sanford.ims.alaska.edu) wrote:]
>: The above is correct. The telco usage of DC on the cable has
>: nothing to do with longitudal balance. It is used often in an
>: unbalanced manner with no effect on the impedance or balance of
>: the line.
>
>I think this is where we have the terminology problem. I am interpreting
>balance to mean longitudinal balance i.e. the relative impedance to
>ground of Tip and Ring. There is another use of balance to refer to
>current flow.
The question of how T/R can be a -48 and 0 volts involves
longitudinal balance only as minor aspect (in that said balance
must not be upset by DC circuits, which therefore require
isolation). The question is how can the DC voltage be at ground
potential if the line is DC balanced, and the answer is, once
again, it is NOT a DC balanced device. It is an AC balanced
device, which requires longitudinal balance. The discussion of
longitudinal balance is interesting, but... longitudinal balance
is not specified at DC, only for AC. It is irrelevant at DC.
Here is a diagram that perhaps helps explain it:
REPEAT COIL
(aka a transformer)
Tip 0-----------) || (--------------------------------//---0
) || (
) || (
) || (
Balanced ) || (-------*------------+
Voice ) || 2 | | Twisted
Frequency ) || mfd === | Pair
Circuit ) || cap | | Cable
) || (-------*-----+ |
) || ( | |
) || ( | |
) || ( | |
Ring 0-----------) || (------------->------>-----------//---0
| |
| |
| |
| |
| |
A Lead <---------------------------^------+
|
DC Circuit |
|
B Lead <---------------------------+
That is exactly the method by which the AC and DC circuits are
separated. The AC circuit _must_ be balanced to provide isolation
from crosstalk, and any unbalance will cause degraded
functionality.
The DC circuit may or may not be balanced and has little effect on
the AC circuit. Generally the DC circuit is not balanced because
it operates in a low resistance current limited mode, hence the
impedance for intefering randon noise is low, and the induced
current is significantly less than the DC signal.
The common way that the above is used for a dial line is that one
side has 48v resistance talk battery applied and the other side
has ground applied. The resistance in series with either side may
be adjusted in some cases to provide reasonable values of loop
current. There is no need for the resistances of both sides to be
equal to maintain balance, longitudinal or otherwise.
One can experimentally prove that by shunting each side of the
AC circuit to ground with two different sizes of resisters. Say
600 ohms on one side and 1200 ohms on the other side. Bingo, bad
circuit! Do the same to the DC circuit it will still function.
Case closed,
Floyd
Jay Ashworth
>2. If we feed ground to T and battery to R through equal resistances we
>will be balanced to ground at DC. If we feed through unequal resistances
>to ground (for example connecting Tip directly to ground) we will be
>unbalanced at DC. I am making no distinction here between AC and DC
HUH?
"Balanced", at DC, referes to "balanced WRT earth ground", does it not?
Assuming we agree on the definition, then a telco station line cannot be
balanced to ground if one wire is above (or below :-) ground, and the
other is at ground.
It is, by definieion, unbalanced to ground.
What did I miss here?
Mike Sandman
This is a question for Wallace...
I've been a phone man for about 25 years, and most of your explanation of
balanced lines was way over my head!
I think of balanced lines as both the tip side and ring side of the line being
the same distance. When they're not, you get assorted noises, clicking and
crosstalk.
When I report a line to the phone comany as being noisy, they often call back
with a "no trouble found" reply. Any person who wasn't deaf could hear the
noise, but that's the fun of dealing with the phone company (any of 'em!).
Can I actually use the circuit below to measure the voltage drop across a
resistor to verify for myself that the line is imbalanced? Does the
diagram show resitors going from the ring and tip, shorted to a 450
Ohm resitor, and then thru S1 to ground?
Thanks in advance... Mike Sandman
>You can measure DC balance with a resitive networks as follows:
> Ring Tip
> < <
> Rr < < Rt
> < <
> |_______|
> |
> <
> < Rg
> <
> |__
> ___ |- s1
> _|_
> / / /
>Rr and Rt should be matched to 0.5% for a resonable measure. (Otherwise
>this network will be unbalanced when connected to ground). Rg can be about
>450 ohms. If you connect this network across tip and ring, the metallic
>voltage across tip and ring will be x where x is a function of the office
>battery and total loop resistance. The line will register off-hook and you
>will receive dialtone. If you momentarily press s1 there will be no change
>in the metallic voltage across Tip and Ring for a line that is resistively
>balanced to ground. There will be a change in the current in the TIP and
>the current in RING unless the battery if floating. The voltage across Rt
>and Rr will also be different unless the battery is floating (or both
>positive and negative batteries with common ground are used (very rare).
>W. A. Ritchie
>Ft. Lauderdale, Florida
Pat
opinions exist as to what is a "balanced circuit." So, how about all you
experts sharing your opinions? I'll try to collect them all, and give
the rest of us a list to choose from. This should be interesting. It
goes like this:
IMHO, a balanced circuit is...