Cryptic Modem Jargon

Cryptic Modem Jargon

Post by r.. » Tue, 07 May 1996 04:00:00



I was wondering whether or not anyone would be kind enough to
translate into english the following modem jargon for me.
Technical-ish replies will be understood.

No doubt this isn't the first time these questions have been posed,
but I can't find anything in the group that covers them.

1)      "long space disconnect"
2)      "Guard Tone"             and what purpose does it serve?
3)      "Pulse make/break ratio" what purpose does it serve?
4)      "Auto-reliable fallback"
5)      "disconnect buffer delay"
6)      "Automatic rate adaption"
7)      "Automatic rate adaption"-low BER    
8)      "Automatic rate adaption"-medium BER  
9)      "Automatic rate adaption"-high BER    
10)     "LAPM with MNP fallback"
11)     "LAPM with MNP normal fallback"

Thanks, in advance, for any help which may be forthcoming.

Siegen

 
 
 

Cryptic Modem Jargon

Post by Geoffrey Wel » Thu, 09 May 1996 04:00:00



>I was wondering whether or not anyone would be kind enough to
>translate into english the following modem jargon for me.
>Technical-ish replies will be understood.

>No doubt this isn't the first time these questions have been posed,
>but I can't find anything in the group that covers them.

I'll give it a shot...

Quote:>1)  "long space disconnect"

I believe that a "long space" is also known as a 'BREAK' signal.  Essentially,
it's a signal that can't be confused for any legal data value on an
asynchronous serial connection (i.e. one with start and stop bits).  I also
believe that some modems can be configured to hang up on receiving it.

Quote:>2)  "Guard Tone"             and what purpose does it serve?

It's just an extra tone played on the phone line so that telecom equipment
knows that there's a data connection under way and won't be tempted to hang
up.

Quote:>3)  "Pulse make/break ratio" what purpose does it serve?

When you pulse dial, you're shorting (well, nearly) the two telephone wires
together ('making' the loop) and un-shorting them ('breaking' the loop) in
order to make a pulse.  The phone system expects a certain ratio of time
shorted and time not shorted, as well as a certain dial speed (approx. 10
pulses per second, though 20 works on some equipment).  The European and North
American phone systems expect a slightly different ratio, though I'd be
surprised if their tolerances weren't enough to overlook an incorrect setting.

Quote:>4)  "Auto-reliable fallback"

The modem will try to establish an error-correcting (LAP-M or MNP) link, but
will stay connected if it doesn't work out.  'Reliable' mode will hang up even
a perfectly good connection if the error correction negotiation doesn't
succeed, e.g. the modem on the other end doesn't support error correction...
or its an RPI modem running without the WinRPI driver.  <grin>

Quote:>5)  "disconnect buffer delay"

When an error-correcting link is established, the modem buffers a fair amount
of data internally.  This causes a problem when, for instance, a BBS sends you
a logout screen and then hangs up; the data that the BBS sent may not have
been transmitted by the time it hangs up, so the caller gets half a screen
ending with the words "NO CARRIER", usually in some odd colour.  <grin>  By
forcing a delay before disconnecting, the modem increases the chances that the
data will actually be sent before it hangs up.

Quote:>6)  "Automatic rate adaption"

The dynamic adjustment of the communications speed to compensate for changing
line conditions as evidenced by data errors.

Quote:>7)  "Automatic rate adaption"-low BER    
>8)  "Automatic rate adaption"-medium BER  
>9)  "Automatic rate adaption"-high BER    

BER = Bit Error Rate (formally; it may be used to represent the Block Error
Rate, though some manufacturers - including USRobotics - use the more formal
symbol "BLER" for that).  These settings probably allow you to tell the modem
how tolerant you want it to be of errors before it changes the data transfer
speed.

Quote:>10) "LAPM with MNP fallback"
>11) "LAPM with MNP normal fallback"

Can I presume that you know that the Link Access Protocol for Modems and
Microcom Networking Protocol are?

#10: If LAP-M negotiation fails, try MNP.  If that fails, hang up. (*)
#11: If LAP-M negotiation fails, try MNP.  If that fails, stay connected with
no error correction protocol.

(*) Some documentation may speak of an MNP fallback _without_ implying that
the alternative is to hang up; i.e. if #11 is not an explicitly mentioned
alternative, you should assume that the words beside your #10 above mean what
I said about #11.

Quote:>Thanks, in advance, for any help which may be forthcoming.

In retrospect, I think that I got most of them near-right.  I got better as I
could see the end looming in the distance.

--
            Geoffrey Welsh, Developer, InSystems Technologies Inc.


            TYPING IN ALL CAPS IS GROUNDS FOR IMMEDIATE DISMISSAL.

 
 
 

Cryptic Modem Jargon

Post by Stephen [kiwin] PA » Mon, 13 May 1996 04:00:00



rod> I was wondering whether or not anyone would be kind enough to
rod> translate into english the following modem jargon for me.
rod> Technical-ish replies will be understood.

rod> 2)      "Guard Tone"             and what purpose does it serve?

In V.34, it is a 1800Hz tone to keep some older switching equipment
from thinking that modem tones are actually in-band switching commands.

rod> 3)      "Pulse make/break ratio" what purpose does it serve?

for pulse dialing (ie, not tone dialing), it is the lengths
of time between clicking on and off.
Different countries have slight different values.

rod> 6)      "Automatic rate adaption"

In V.34, once the modem has trained up, the data rate can been
changed (by the modems) to account for changes in line conditions
or the number of errors.  In Rockwell DataPumps, ARA also
helps determine the initial connect rate.  It is typically
based on Eye Quality Monitor (EQM).

rod> 7)      "Automatic rate adaption"-low BER    
rod> 8)      "Automatic rate adaption"-medium BER  
rod> 9)      "Automatic rate adaption"-high BER    

BER = Bit Error Rate.
When there is low BER, that means not many errors are being received.
With high BER, the modems would likely negotiate a lower data
rate to hopefully reduce the number of errors occuring.
There is a tradoff between having a higher percentage of errors
and a higher data rate (the errored framed must be resent)
than to fallback to a lower data rate and have less errors.

rod> 10)     "LAPM with MNP fallback"
rod> 11)     "LAPM with MNP normal fallback"

Both are error correction protocols. Those statements reflect some
giving preference of one protocol over the other during the negotiation
phase.

You may want to look over the Rockwell Generic AT command manual at:
   http://www.nb.rockwell.com/ref/reference.html

regards, kiwin
-----
Stephen [kiwin] Palm                  COMNET 930-1564      T: +81-3-5371-1564
Rockwell Semiconductor Systems       (JST=PST+17hours)     F: +81-3-5371-1507


--
-----
Stephen [kiwin] Palm                  COMNET 930-1564      T: +81-3-5371-1564
Rockwell Semiconductor Systems       (JST=PST+17hours)     F: +81-3-5371-1507


 
 
 

Cryptic Modem Jargon

Post by Christian Weisgerb » Mon, 13 May 1996 04:00:00



> >1)     "long space disconnect"

> I believe that a "long space" is also known as a 'BREAK' signal.  Essentially,
> it's a signal that can't be confused for any legal data value on an
> asynchronous serial connection (i.e. one with start and stop bits).  I also
> believe that some modems can be configured to hang up on receiving it.

Yes, "long space" is a (long) BREAK.

Back in the days before error control was commonplace, when one modem
hung up the other modem would take some time to detect the loss of
carrier and meanwhile "recognize" and send garbage to the host. To
circumvent this, a simple hang-up protocol was conceived: before
actually* up, a modem would send a long BREAK. On reception, the
other modem would clamp its receiver.
The sending modem transmits 4s of continuous BREAK before* up,
the other modem needs to receive one for at least 1.5s.

Normal BREAK signals used for various purposes are much shorter. 0.25s
is a common value.

Quote:> >2)     "Guard Tone"             and what purpose does it serve?

> It's just an extra tone played on the phone line so that telecom equipment
> knows that there's a data connection under way and won't be tempted to hang
> up.

From a posting by Toby Nixon in 1992:

--------------->
The guard tone is for an entirely different purpose.  In certain
older telephone systems, "in-band signalling" (tones) is used to
indicate when the call is to be disconnected; when you hang up your
phone, the telephone switch sends the disconnect tone to the other
switch to clear the call.  Nowdays, the vast majority of telephone
networks use "out-of-band" signalling -- completely separate
communication paths between switches for call control -- and tones
sent on the line have no affect on the call.  But on these older
networks, if the modem happened to produce a signal that sounded
enough like the disconnect tone, it would cause the switch to
disconnect the call -- quite disconcerting!  In particular, V.22 and
V.22bis modems are capable of producing such tones.  Fortunately,
these older switches are designed so that they recognize the
disconnect tone only when it is sent pretty much by itself; if
signals are present in other parts of the frequency band, the tone
is not recognized.  The guard tone (either 550 Hz or 1800 Hz) is
sent continuously during the call by the answering mode,
specifically to provide this "extra tone" and thereby defeat the
recognition of the disconnect signal in the rare event that the
modem would produce it spontaneously.  And, as I said, the nearly
complete replacement of old in-band signalling systems with
out-of-band signalling systems is making guard tones unnecessary.
<---------------

--

  See another pointless homepage at <URL:http://www.veryComputer.com/~naddy/>.

 
 
 

Cryptic Modem Jargon

Post by Christian Weisgerb » Sun, 19 May 1996 04:00:00



Quote:> rod> 2) "Guard Tone"             and what purpose does it serve?
> In V.34, it is a 1800Hz tone to keep some older switching equipment
> from thinking that modem tones are actually in-band switching commands.

There is no guard tone in V.34, it was only used with V.22/V.22bis.
In fact, adding a 1800Hz tone looks to me like a good way to kill a V.34
connection.

--

  See another pointless homepage at <URL:http://home.pages.de/~naddy/>.

 
 
 

Cryptic Modem Jargon

Post by Stephen [kiwin] PA » Sun, 19 May 1996 04:00:00



Quote:> rod> 2) "Guard Tone"             and what purpose does it serve?
> In V.34, it is a 1800Hz tone to keep some older switching equipment
> from thinking that modem tones are actually in-band switching commands.

Naddy> There is no guard tone in V.34, it was only used with
Naddy> V.22/V.22bis.  In fact, adding a 1800Hz tone looks to me like a
Naddy> good way to kill a V.34 connection.

Sorry for not being explicit there.

The Guard Tone is used in the V.34 training (startup) and the Control
Channel of the half duplex mode of V.34.  That control channel looks
an awful lot like V.22/V.22bis.

See 10.1.2.1/V.34 , 10.1.2.3.1/V.34, 10.2.4/V.34.

You are correct that it is not used during the high speed data portion.

regards, kiwin
--
-----
Stephen [kiwin] Palm                  COMNET 930-1564      T: +81-3-5371-1564
Rockwell Semiconductor Systems       (JST=PST+17hours)     F: +81-3-5371-1507


 
 
 

1. Cryptic Abbreviations

Patrick, I don't think it would be such a bad idea to put out a
glossary of telecom terms to stay on here all the time. I get lost
myself when we start talking about all those cryptic acronyms.

/*  John Wheeler - Unix/C Systems Designer/Programmer/Administrator/etc...    *
 * Turner Entertainment Networks * Superstation TBS * TNT * Turner Production *
 *           ...!gatech!nanovx!techwood!johnw    (404) TBS-1421               *
 * "the opinions expressed in this program are not necessarily those of TBS"  */

[Moderator's Note: Well, the beginnings of the glossary are under
construction now. I am trying to get a file transferred to me with
quite a few terms defined.  Stick around for a few days; I should have
it installed.  PT]

2. TAO and Orbix

3. ReL Cryptic Abbreviations

4. Is there a winsock-compliant PPP with CHAP?

5. Cryptic Abbreviations

6. Atari ST sound-player: lps1.21 uploaded

7. Cryptic 4.11 Install Error?!?!

8. Consultants

9. Holy Change Agent! Consultants Edit Out Jargon

10. Jargon Overload

11. When Technical Jargon Requires Hyphens

12. Jargon Dictionary Wanted