NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/dcom/modems-part1/9470

Internet Message Format | 1996-08-05 | 6.2 KB

Path: polarnet.com!floyd
From: [e-mail] (Floyd Davidson)
Newsgroups: comp.dcom.modems,comp.dcom.telecom.tech
Subject: Re: Making a RJ11 "busy" plug
Date: 28 Mar 1996 09:11:48 GMT
Organization: __________
Message-ID: <4jdl4k$[e-mail]>
References: <4i4n3j$[e-mail]> <4j20vk$[e-mail]> <[e-mail]> <[e-mail]>
Reply-To: [e-mail]
NNTP-Posting-Host: tanana.polarnet.com

[e-mail] wrote:
>Central office switches use a current source to feed the
>loop current. This current source is defined by the
>following equivalent circuit:
>
>
>   ----------------------/\/\/\-------------/\/\/\---------|
>   |                    400 Ohm            Loop Resistance |
>   |              Battery Feed Resistor     25 Ohm min     \
>   |                                      1300 Ohm max     / Telephone
> ------  48 Volt DC                                        \ Set
>   --   Constant Voltage                                   /
>    |     Source                                           \
>   |                                                       |
>   |                                                       |
>   ---------------------------------------------------------
>

The "48 Volt DC Constant Voltage Source" invariably will be 52 VDC,
(and when not invariable, it can be as high as 56 VDC and as low as
46 VDC).

>The telephone set has a nominal resistance of about 200 Ohms.
>However, it must be current limited to a maximum current of 90 mA.
>It also must be able to work with loop currents as low as 23 mA.

An off hook condition is defined as 20 ma in the US.  The
difference is interesting (and contributes to reliable operation).

>Since a modern switch is built with a *current* source, it
>can detect excessive loop current and turn the loop down or
>generate a trouble message for shorted loops.

Not necessarily true.  It could detect excessive current, but in
fact I know of not one that does.  What they do is detect the
difference between 20 ma, more or less.  That is, specifically
they detect on/off hook.  I am most familiar with Nortel switching
units and a huge variety of FX Station cards for Channel Banks
(Though I worked on a real live Stroeger 50B for a couple years
once upon a time!), and have never heard of a switch that does in
fact detect "excessive loop current".  Actually I've never seen a
switch that can even deliver excessive loop current, but...

What is generated however, is a "Permanent Signal" alarm.  That is
old terminology used with mechanical switching systems, and for
any given type of common control switch it might have a different
name for whatever type of report is generated.  A PS alarm means
the line has been off hook with no normal action (such as dialing)
having taken place within a preset time limit.

An individual Permanent Signal alarm is universally ignored other
than for the automatic intercept-announcement / howler / out-of-
service routine we have all experienced with phones left off hook
for too long.  As a group they might well be counted for stats.
No manual intervention is normally initiated due to an individual
PS alarm (because doing so would double the labor cost of a telco!).

>Ohms law tells you that the set's resistance can be between 350 ohms
>for a short loop, and about 50 ohms for a long loop.  If you are
>on a short loop, and you exceed 90 mA, the current source in the switch
>can get very upset.  The specification for telephone instruments,
>EIA/TIA-470-A, states that the instrument must limit current at 90 mA.

I do not have access to a copy of the EIA specs.  I have heard
from others that it says that 240 ohms is the maximum resistance
guaranteed to cause an off hook condition (i.e., to allow 20 ma of
current).  What I would like to know is exactly what it states in
regard to current limiting.  Telephone sets have traditionally not
had the ability to limit current in the loop, but have always had
the ability to limit current through the set.  That limitation
used to be provided by a varister across the line which would
limit the voltage appearing across the teleset.

>If you are on a short loop, and you use a short circuit to make your
>"busy" plug, you could exceed this current.  In my experience I have
>never had any problem using a "short", but I have never been
>that close to the switch either.

I've never had a problem either, and I've never been far from the
switch.

>If you use a 100 Ohm resistor, it should be rated at one-watt.

That 100 ohm resistor, using the 90 ma loop current figure, could
be called upon to actually dissipate .81 watts, and therefore should
have a rating for at least 2-4 times that value.  For example, at
the same current it will dissipate twice the heat that the 47 ohm
resistor mentioned in the following paragraph would, but that
resistor has twice the rating!  To be just as conservative, a 4
watt 100 ohm resistor should be used.

>Older modems without a solid-state current-sink generally
>used a 47 Ohm, 2 Watt resistor in their DAA.  Newer modems
>have a transistor current-sink set at 25-35 mA.
>
>Some cheap PBX's apparently under-size power rating of
>the current sink in their trunk cards.  Mike Sandman devotes two
>full pages in his catalog to loop current limiting devices.
>He wouldn't be selling these things if people weren't toasting
>their trunk cards.

One definite caveat in heeding all of us old telephone types, is
we work on top quality equipment and rely on it.  Consumer quality
key systems and small PBX's may not have that quality.  While I
will stand by the "use a shorting plug" 100% when it comes to how
to treat the telco's lines, I am not not willing to go that far if
you own a small PBX or key system.  (I also would highly recommend
avoiding any system that can't handle a shorted loop, right at the
card, for a year or two without damage...)

Otherwise, I'll just repeat that any telco tech seen using a
resistor of any size or power rating to cause an off hook would
be chastized.

Floyd
--
Floyd L. Davidson          Salcha, Alaska         [e-mail]