[comp.dcom.modems] Will T2500 cycle all speeds if S50=3?

jim@crom2.uucp (James P. H. Fuller) (03/18/91)

     Because several of my local callers have old and/or dimwitted modems
that get confused and can't connect if I set my T2500's S50 register to 0
(automatic speed determination, with either PEP or V.32 tones first,
depending on S92) I presently have S50=3.  That sends 2400bps tones first
and the local folx are happy, but as I read the manual (it's not totally
clear on this point -- or maybe it is but I just don't like what I'm read-
ing...) the modem is ONLY going to cycle among 2400-1200-300bps and NEVER
send PEP or V.32 tones.  Is this correct?  Is there a register setting that
will start with 2400bps tones but still cycle through ALL the others and
give a fast connect to somebody calling in PEP or V.32 mode?  The reason
this has suddenly become significant is that I have groff 1.0 binaries for
ISC Unix that I want to make available to the world in general via anon-uucp
and while nobody around here ever calls in PEP mode I expect that people
looking for groff would want a high-speed connection only (it's not as big
as emacs but it's still a whopper.)
                                                Thanks
                                                James P. H. Fuller
                                                jim%crom2@nstar.rn.com

davem@nro.cs.athabascau.ca (Dave McCrady) (03/20/91)

jim@crom2.uucp (James P. H. Fuller) writes:

> 
>      Because several of my local callers have old and/or dimwitted modems
> that get confused and can't connect if I set my T2500's S50 register to 0
> (automatic speed determination, with either PEP or V.32 tones first,
> depending on S92) I presently have S50=3.  That sends 2400bps tones first
> and the local folx are happy, but as I read the manual (it's not totally
> clear on this point -- or maybe it is but I just don't like what I'm read-
> ing...) the modem is ONLY going to cycle among 2400-1200-300bps and NEVER
> send PEP or V.32 tones.  Is this correct?  Is there a register setting that
> will start with 2400bps tones but still cycle through ALL the others and
> give a fast connect to somebody calling in PEP or V.32 mode?  The reason
> this has suddenly become significant is that I have groff 1.0 binaries for
> ISC Unix that I want to make available to the world in general via anon-uucp
> and while nobody around here ever calls in PEP mode I expect that people
> looking for groff would want a high-speed connection only (it's not as big
> as emacs but it's still a whopper.)

  S50=3 forces 2400 bps ONLY.  Connects at any other rate will fail.
You should set S50=0. Then set S92=1.  S92 controls the sequence of 
answer tones issued by the modem (pp 56-2 and 56-3 of the manual).


   Dave McCrady                 |  ersys!davem@nro.cs.athabascau.ca
+1 403-454-4054 (voice)         |  Disclaimer? Who needs a disclaimer?
+1 403-454-6093 (public USENET) |  I >own< the joint.

gandrews@netcom.COM (Greg Andrews) (03/20/91)

In article <1991Mar18.143749.1250@crom2.uucp> jim@crom2.uucp (James P. H. Fuller) writes:
>
>     Because several of my local callers have old and/or dimwitted modems
>that get confused and can't connect if I set my T2500's S50 register to 0
>(automatic speed determination, with either PEP or V.32 tones first,
>depending on S92) I presently have S50=3.  That sends 2400bps tones first
>and the local folx are happy, but as I read the manual (it's not totally
>clear on this point -- or maybe it is but I just don't like what I'm read-
>ing...) the modem is ONLY going to cycle among 2400-1200-300bps and NEVER
>send PEP or V.32 tones.  Is this correct?  
>

Yes, that's correct.  Setting S50=6-1 limits the modem to that connection
speed or slower speeds (if any) when fallback is enabled with S94=1.

>
>Is there a register setting that will start with 2400bps tones but still 
>cycle through ALL the others and give a fast connect to somebody calling 
>in PEP or V.32 mode?
>

No, there isn't.  The best that you can do is to set S50=0, S92=1, and
perhaps adjust S93 to a smaller value.  S93 determines the length of the
"AC" tone used by V.32 modems, so be careful.  The more you reduce the
"AC" tone, the lower your chances for getting V.32 connects.

There are still come 1200 and 2400 modems out there that barf big time
on the V.32 "AC" tone.  There's very little you can do to the answer
modem to compensate for the calling modem in this case.  The caller
could keep their modem off-line during the V.32 tones, and kick it
on-line afterward, so the modem doesn't see the tone it can't handle.
Another possibility would be for the caller to increase the values of
S9 and S10 in their modem when they call.  Neither of these methods
is a sure-fire fix, but that's all I can come up with at the moment.

Another possibility would be to somehow set your system up to answer
2400-only at certain times of day (S50=3).  The rest of the time, it 
would allow all speeds (S50=0, perhaps with S92=1), and users of old
1200/2400 modems would have to take their chances during that time.

Hope this helps...

-- 
.------------------------------------------------------------------------.
|  Greg Andrews   |       UUCP: {apple,amdahl,claris}!netcom!gandrews    |
|                 |   Internet: gandrews@netcom.COM                      |
`------------------------------------------------------------------------'

davidg%aegis.or.jp@kyoto-u.ac.jp (Dave McLane) (03/20/91)

jim@crom2.uucp (James P. H. Fuller) writes:

>      Because several of my local callers have old and/or dimwitted modems
> that get confused and can't connect if I set my T2500's S50 register to 0
> (automatic speed determination, with either PEP or V.32 tones first,
> depending on S92) I presently have S50=3.  That sends 2400bps tones first
> and the local folx are happy, but as I read the manual (it's not totally
> clear on this point -- or maybe it is but I just don't like what I'm read-
> ing...) the modem is ONLY going to cycle among 2400-1200-300bps and NEVER
> send PEP or V.32 tones.  Is this correct?

Sort of.... but not quite.

According to the manual + my experience, there are four registers
controller this function: S50 controls the maximum/only speed
depending on S94; S92 controls whether the PEP tones are presented
first or last (assuming they are included in those allowed by S50;
S90 is used to determine whether to use Bell 103/V21 mode.  This is
not new information, it's in the manual; I'm only saying that I
have found it to be the case.

Note that S92 controls *only* the presentation of the PEP answer
tones; if you run S50=0 and S92=0 the answer tones are PEP, 9600,
2400, 1200, 300; if you run S50=0 and S92=1 the answer tones are
9600, 2400, 1200, 300 PEP. In other words the 9600 tones stay in
the same place.

You *can* put a 9600 tone in front with S92=2 but I suspect that
the (unspecified) confusion (like what do they do when the are
confused?) may be coming from the 9600 answer tones.

When I changed the modem on Aegis to the T2500 some people had to
change the value in their modems/scripts that waited for a CONNECT
or the prompt as due to the increased time due to having to wait
through the 9600 answers tones. The confusion you mention couln't
be this, could it?

--Dave

grahj@gagme.chi.il.us (jim graham) (03/21/91)

In article <PFP5y1w163w@aegis.or.jp> davidg%aegis.or.jp@kyoto-u.ac.jp (Dave McLane) writes:
>jim@crom2.uucp (James P. H. Fuller) writes:
>
>>      Because several of my local callers have old and/or dimwitted modems
>>  [ and so on....this has been copied 10,000 times....you get the idea ]
>
>According to the manual + my experience, there are four registers
>controller this function[.]
>
>Note that S92 controls *only* the presentation of the PEP answer
>tones; if you run S50=0 and S92=0 the answer tones are PEP, 9600,
>2400, 1200, 300; if you run S50=0 and S92=1 the answer tones are
>9600, 2400, 1200, 300 PEP. In other words the 9600 tones stay in
>the same place.
>
>When I changed the modem on Aegis to the T2500 some people had to
>change the value in their modems/scripts that waited for a CONNECT
>or the prompt as due to the increased time due to having to wait
>through the 9600 answers tones. The confusion you mention couln't
>be this, could it?

thanks --- you saved me from digging through the manual myself to answer
this one....  there is quite definitely a way to make the T2500 cycle 
through the lower speed stuff, and then do PEP last.  several of the
systems I call have their modem setup this way.  when calling them, I
set MY S50=255 for PEP only (my dialer does this automagically).  the
modem then waits until it hears the PEP answer tones.

I don't mean this to be a RTFM --- please don't take it that way, but 
the documentation for the T2500 is really quite good (excellent, in 
fact!!) --- when you read the descriptions on registers, be sure to follow
the pointers to other related registers....you'd be surprised at what you 
find.

  --jim

------------------------------------------------------------------------------
Share and Enjoy!  (Sirius Cybernetics Corporation, complaints division)
73, de n5ial 

Amateur Radio:
     TCP/IP:    jim@n5ial.ampr.org --- 44.72.47.193
     Packet:    n5ial@wb9mjn (Chicago, IL USA) 
Internet:  grahj@gagme.chi.il.us
------------------------------------------------------------------------------