Index Home About
From: John Nagle <decwrl!well.sf.ca.us!well!nagle@uunet.uu.net>
Subject: Re: Help Needed Understanding ISDN
Date: 9 May 91 07:09:07 GMT

csense!bote@uunet.uu.net (John Boteler) writes:

> Unless the power fails at your location. Then, no more ISDN.

> OOPS!

     No, ISDN sets can be powered from the phone line.  The power
situation is ingenious.  Normally, you can draw (I think) 400ma
off-hook from an ISDN line.  This should be enough for a reasonable
phone, and maybe a digital answering machine as well.  If the CO has a
power problem (maybe when commercial power is out) the DC polarity of
the line is reversed, and you can then draw only some lesser amount of
power.  Maybe your dial light will go out.  But the phone should still
work.

     Now, which ISDN phones properly comply with the spec?


John Nagle

From: fgoldstein@bbn.|nospam.|com (Fred R. Goldstein)
Newsgroups: comp.dcom.isdn
Subject: Re: 3Com Impact IQ: Data-Over-Voice / LEDs
Date: 27 Jun 1997 13:38:32 GMT

In article <5ovbmm$omi@usenet1.interramp.com>, Dan@dan.com says...

>They are NOT the same. It appears your ISP allows 56k DATA calls on the
>so-called DOV DN. The ISP may also be configured to accept analog modem
>connections. Try this test:
>        1. Dial the DATA DN using a telephone connected to one of the voice
>ports. Do you hear what sounds like a distorted tone (clipped sine)? If so,
>then it's flags - a DATA call.

Whoa.  If you dial a "Data" number from an analog phone, one of three things
could occur:

1)  It is rejected (telco recording/tone) because it's incompatible.
2)  It answers with a regular modem tone sequence, because it uses Bearer
Capability to distinguish modems from ISDN data.  This is the norm, for
instance, on UUNET dialins, which are MAXes that don't sort by number.
3)  It answers as a data call anyway, because some COs (notably DMS) will not
take reject for an answer and improperly treat a data-only number as a voice
number when called from a voice phone.  In this case a data call is usually
almost silent, punctuated by ticking from packets.

>        2. Dial the DOV DN the same way. Do you hear the same distorted tone
>or do you hear a modem answer tone (clean sounding tone)?

A DOV number should answer with a LOUD 1000 Hz tone, perhaps what was called
"clipped" because it's not a sine wave.  That's because it's 56k, not 64k.

In any case, the audible behavior of the destination has nothing to do with
billing.  Also I suspect that PortMasters do some tricks that enable a the
same number to answer with modems or DOV, so they don't return the 1000 Hz
tone until they hear it.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
BBN Corp., Cambridge MA  USA         +1 617 873 3850
Opinions are mine alone; sharing requires permission.


From: fgoldstein@bbn.|nospam.|com (Fred R. Goldstein)
Newsgroups: comp.dcom.isdn
Subject: Re: 3Com Impact IQ: Data-Over-Voice / LEDs
Date: 30 Jun 1997 02:55:17 GMT

In article <5p1ii2$7ku@usenet1.interramp.com>, Dan_Brennan@3mail.3com.3com
says...

>Fred,
>I agree, I missed your first point. If the destination is definitely without
>a doubt clear channel ISDN DATA only, then the call will not connect!

I know of few numbers that are provisioned ISDN DATA only, though they exist
("Termtype=E", in old AT&Tese) on the 5E.  A DMS can't do it, though a TE
could reject non-data calls.

I was referring to the inverse, a data-bearer call to a voice-only number.
It should not connect.  You know that, I know that, the ITU knows that.
However, somebody who wrote some DMS100 code didn't know that!  A data bearer
call to a voice-only number will get, I think, a PROGRESS message (not sure
of the coding) and continue; the receiving end will receive a "voice" call.

Just in case that was what I was unclear about.  (BTW I was quite surprised
when that was noted here; it might be a NYNEXism, for all I know.)
--
Fred R. Goldstein   k1io   fgoldstein@bbn.com   +1 617 873 3850
Opinions are mine alone; sharing requires permission.


From: fgoldstein@bbn.|nospam.|com (Fred R. Goldstein)
Newsgroups: comp.dcom.isdn,panix.isdn
Subject: Re: ISDN Router that supports bonding (112K) DOVBS?
Date: 30 Jun 1997 16:00:49 GMT

In article <5p4795$jmf@panix.com>, tls@panix.com says...

>The point is that some people say "voice" when they mean "speech", and some
>people say "voice" when they mean "3.1KHz audio".  There's a distinction, and
>the use of a term that doesn't specifically mean either causes quite a bit of
>confusion.  For example, it can lead naive ISDN customers here to use the
>"speech" bearer capability for their calls to Panix, when they should
>definitely not use that -- they should use the "3.1KHz audio" BC.  "Speech"
>calls may be routed over analog or compressing facilities, or otherwise
>recoded almost arbitrarily...

I've taken to using "voice" for precisely the reason you refer to.  There is
no "voice" BS, but if you specify *either* 3.1 kHz audio .OR. speech, the
tariff treatment is the same.  In practice, NOBODY (at least in the USA)
implements the two separately; "audio" is the default, and "speech" is taken
to be an oddball encoding (in DNS terms, think of a CNAME) of same.  Yes, I
know very well what Q.931 says, but that was written a LONG time ago.  David
Lesher's point about "DO3.1kHzAudioBS" is well taken; in practice,
"3.1kHzAudio" is spelled and pronounced the same as "voice".

>I said, "speech calls with an A-law encoding field".  If the gateway tandem
>thinks it's already *got* G.711 A-law, it's not supposed to try to turn it
>into A-law again. :-)  The question is whether your call will _get_ to the
>gateway at all; it should, but I wouldn't necessarily count on it.

Good question.  My suspicion is that a call already coded A-law will get
bounced by a gringo switch, if the switch even looks at that codepoint.  But
even if you got to the gateway intact, it'd fail on an overseas call because
undersea cable "voice" bandwidth is usually compressed, at least to 32 kbps
ADPCM.  I'd be interested to hear if anyone's going uncompressed.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
BBN Corp., Cambridge MA  USA         +1 617 873 3850
Opinions are mine alone; sharing requires permission.



From: fgoldstein@bbn.NO$LUNCHMEAT.com (Fred R. Goldstein)
Newsgroups: comp.dcom.isdn
Subject: Re: NT-1 & NT-2 What exactly is the difference
Date: 13 Aug 1997 13:52:49 GMT

In article <5srsok$27a$1@plug.news.pipex.net>, fdriver@rnib.org.uk says...

>I am trying to teach myself a bit more about ISDN (a subject I only
>know a little about).  However I am getting conflicting information on
>exactly what the difference is between NT1 and NT2.

NT1 is a line terminator, like a CSU.  It terminates the Digital Section
(local loop from subscriber to telco CO) and provides the subscriber with a
4-wire "S/T" short-span interface.  It only operates at Layer 1.

NT2 is ISDN-ese for "PBX", or in theory any other device that actually gets
invovled with call control signaling ("layer 3" of the control plane) but is
not the terminal-device.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
BBN Corp., Cambridge MA  USA         +1 617 873 3850
Opinions are mine alone; sharing requires permission.




From: fgoldstein@wn.do-not-spam-me.net (Fred Goldstein)
Newsgroups: comp.dcom.telecom.tech
Subject: Re: If phones were built new today?
Date: Tue, 19 Jan 1999 03:41:40 GMT

On 18 Jan 1999 17:01:14 GMT, hancock4@bbs.cpcn.com (Lisa or Jeff)
wrote:

>[heavy expense of wiring]
>> So please explain how using a local battery system was going
>> reduce that cost?
>
>At present, you are transmitting relatively high amounts of power
>over a telephonne line for ringing and supervision.  You really
>only need much less amounts of power to carry a conversation
>and signalling accomplished using electronic means instead of
>through electrical.
>
>By using power supplied at the telephone instead of the C.O.,
>you eliminate the need to carry that power over phone lines.

The power sent over the wires is rather miniscule.  The thousands of
dollars needed to get wire to a new location (that was after all the
question) has nothing to do with power; it has to do with the cost of
laying anything at all into a new trench or putting up poles.  Glass,
powered copper, unpowered copper, whatever, it's mostly labor.

My vote:  It'd look somewhat like Euro-ISDN.  All digital, with
sealing current in the local loop plus enough CO power to run a single
digital phone in a basic POTS-like mode.  Extra features can be
locally powered, with batteries if you like, but a little bit of CO
power for emergencies.

American telcos took the Mr. Too-Cheapo approach, and don't supply any
power for  ISDN phones.
--
Fred R. Goldstein   k1io   fgoldstein"at" wn.net
Opinions are mine alone; sharing requires permission.



Newsgroups: comp.dcom.isdn
Subject: Re: Monitoring ISDN connection
From: fgoldstein@bbn.NO$LUNCHMEAT.com (Fred R. Goldstein)
Date: Mon, 14 Sep 1998 14:17:01 GMT

In article <35fb7b83.3333093@news.hal-pc.org>, wdg@[204.52.135.1] says...

>The quality of an ISDN connection does not deteriorate, period. Please
>note the period.
>
>The connect speed of an ISDN connection does not vary or rate-shift. The
>speed you connect at is the speed you stay at, period. Again, please note
>the period.

The bit rate does not change, but the bit error rate does, and thus the
throughput does.  There are lots of variables that can clobber a BRI.  Bad loop
is obvious.  Another is when you have a T1 span with the BR1TE card in the
line, as in SLCs or "virtual ISDN" remote COs.  A misconfigured or
overheating BR1TE card can really through you off, as can a misconfigured DACS
between the originating CO and the BR1TE.  One nasty phenomenon is a
"millihertz" oscillation, wherein every hour or two the line goes up and down,
fading through high BER in the interim.  Have fun fixing that one, which I've
suffered with many a time with the New England Telephone and Telegraph Company
under its various owners.

I'm quite disappointed in the lack of metrics in a lot of ISDN gear, which
seems to assume that the line is always good or else completely out.  Feh!
Remember, most telco repair types go away the second they draw "dial tone";
they are unlikely to do medium-term (say, two hours) BER runs over a BRI.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
GTE Internetworking - BBN Technologies, Cambridge MA USA  +1 617 873 3850
Opinions are mine alone; sharing requires permission.



Newsgroups: comp.dcom.modems,comp.dcom.xdsl,comp.dcom.isdn
Subject: Re: xDSL AND Fibre
From: fgoldstein@bbn.NO$LUNCHMEAT.com (Fred R. Goldstein)
Date: Wed, 10 Jun 1998 18:50:33 GMT

In article <6ll92c$1gt$1@supernews.com>, cgowder@vsta.com says...

fg>>Nice story. Not true, though.  Ma Bell in the olden days ignored data to an
>>imprudent degree.  The Baby Bells (nowadays the Juvenile Delinquent Bells)
>>viewed ISDN mainly as a way to get Centrex features competitive with PBX
>>phones.
>I'm talking about why the underlying design, not necessarly what
>happenen. The Bell system did not do a good job on data
>But that doesn't mean that it was not recognized from an attempt
>to design a future system from the view point at the time
>The thought then right or wrong was one system. Admitly I was
>down in the trenches trying to deploy systems but Have spent some
>time talking to people on the reasons why. I'm also not at all
>trying to defend actual happenings. I'm trying to explain design
>plans in the most simplistic ways I know how.

There were guys in the Labs with all sorts of ideas, some of which were viewed
as wacko by the suits.  By 1980, it was known that ISDN was going to be
developed at CCITT, but AT&T/Bell was behind the curve.  They had begun to roll
out 4ESS toll switches but the 5E was in a nasty quagmire, returned for
regrooving after a disastrous first pass hybrid digital/crossbar (GDX) design.
Their PBX flagship, Dimension, was analog; Antelope (Sys85) was designed but on
hold pending deregulation.

So while some of the techies may well have wanted to design a data-friendly
network, they were not the ones whose ideas became dominant, especially after
Divestiture and the new RBOC management came into being.

>Again to be more specific the first goal I sat in a meeting and listened
>to was 18 kft from a central office and 9 to 15 kf from DLC sights
>because of design limits of the DLC card. Yes the practical length of
>ISDN goes farther. I'm trying to explain an overall plan for deployment
>and keep it simple. You also said the goal was to reach 18 kf for ISDN
>was it reached ?  All deployment meetings I have sat in use distances
>much different but are also strongly driven by other factors than
>transmission.

It sounds like you're referring to the pre-1985 stage.  Back then, AT&T
preferred AMI for the BRI, which had a 12-15 kf practical range.  But by 1985
the RBOCs settled on demanding 18kf, and that's what T1D1.3 demanded, which is
what led to 2B1Q (because NorTel and AT&T/WeCo didn't want Siemens' 4B3T, which
Bellcore favored, to win).

Lots of ideas were batted around in the early days.  It was the deployment
by the RBOCs that went haywire, and continues that way.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
GTE Internetworking - BBN Technologies, Cambridge MA USA  +1 617 873 3850
Opinions are mine alone; sharing requires permission.


Newsgroups: comp.dcom.isdn
Subject: Re: Difference between v.110 vs v.120
From: fgoldstein@bbn.NO$LUNCHMEAT.com (Fred R. Goldstein)
Date: Mon, 27 Apr 1998 19:04:43 GMT

In article <6i29ff$p12$1@shell4.ba.best.com>, megazone@megazone.org says...
>
>Kevin@cyber-wizard.com shaped the electrons to say:
>>The ISP is using equipment from Lucent ... afraind I don't know the
>>model.  I have a Courier I-Modem.
>
>That'd be odd, as the PM-3 doesn't do V.110 normally.  You need different
>DSP cards (the old cards) to do it.  Only a few sites still use them
>while Lucent works on V.110 for the new cards, and all of those are in
>Europe AFAIK.

Sure it does V.110.  It does DOV.  While the V.110 spec includes a lot of
weird and complex things, including an 80-bit frame, those top out at 38.4 or
even 19.2 depending on which version.  And yes, they're pretty much
limited to Europe.  But note:  There is ALSO in V.110 a 56k speed.  It
consists of 7-out-of-8-bit octet stuffing.  This is exactly what the norm is
for 56k data, AND for DOV.

I point this out because DOV uses the 3.1 kHz audio bearer capability exactly
like a modem does, except that it modulates using V.110 at 56 kbps!  This is
not quite the same as the newer, more complex V.90, but it's none of telcos'
business which modem protocol you choose!  DOV is the "V.110 56k modem" over
ISDN.   (V.90 allows one analog loop; V.110 requires end-to-end digital.)

Now V.120 is a whole different animal, rarely used with IP, but sometimes run
inside V.110/56k.
--
Fred R. Goldstein   k1io    fgoldstein"at"bbn.com
GTE Internetworking - BBN Technologies, Cambridge MA USA  +1 617 873 3850
Opinions are mine alone; sharing requires permission.




Index Home About