On Mar 04, 1996 at 16:40, Peter Collis of 3:713/914.16 wrote:

 PC> Sure looks like it works, don't tell Austel they might make Netcomm
 PC> remove the #T command :-)

After stuffing around with tests (to see whether or not it works, so I have
a "solution" for future reference), I've gone back to %L11 (ie.,
-11dBm).

 db>> One assumes that Austel reckons you should only transmit between
 db>> -11dBm and -19dBm, yes?  One makes that assumption since they're
 db>> the only documented settings for the M34F.
 PC>
 PC> I think -10dBm is the limit on loudness can't comment on the other
 PC> direction.

The reason why I picked -9dBm for testing was because I'd seen it mentioned
in a PC telecommunications book that on average, it was "the
best" possible transmit level (in the U.S., anyway).  I've also tried
-10dBm and it didn't help, so it's one of those "just over the
edge" modifications.

 db>> If so, then would it be the USR "at fault" for not accepting
 db>> such connections at a higher speed, or is Telstra proving poor
 db>> quality lines?
 PC>
 PC> Not sure here, USR's seem to come into their own on bad lines, they'll
 PC> work where others give up and drop the line.

Bill seems to think this has something to do with the 64-state trellis
coding;  can someone explain to me *why* this is so?  I have diagrams of
the trellis coding procedure used in V.34, but no rationale.

 db>> (still have a Series 3 M7F in a box here!):
 PC> Power user :-)

You bet!  Although I'm not *that* "feature-centric" that I'll be
getting one of those new 288D modems (I ignored the "Digital
Secretary", too).  I'll wait for V.34bis.

 PC> USRobotics seem to have put a lot of R&D into their DSP based modems
 PC> so maybe Netcomm didn't have the resources to develop something like
 PC> that, i'm assuming the corporate size of USR is far bigger than
 PC> Netcomm, Wonder if USRobotics has a marble foyer :-)

NetComm has been entering into alliances left, right and centre;  they've
just joined forces (Feb 23, 1996) with Boca Research in the U.S. (another
modem maker), so who knows...

 db>> 2.  The problems Paul Edwards is having with two particular
 db>> modems - a V.34 USR Sportster, and a V.32bis (V.32?) Maestro.
 PC>
 PC> I think a lot of Pauls problems are caused by the Qty of his phone
 PC> lines.

Apparently the Maestro SE just died, so maybe *it* was on the way out.  :-)

Cheers..

    - dave
    d.begley{at}ieee.org

---