[pct-l] SHARP TM-20?

[calliger@infolane.com (R.J.Calliger)]

At 8:04 AM -0800 3/5/01, Brick Robbins wrote:
>Not true in this case.
>
>Your voice is analog data. The phone seems to work with it OK. <g>
>Not true in this case.


What I said is true.  I am *very* surprised you questioned me!!

Your voice does not switch at 28KBS or 56KBS or use Kflex!! <g>....

I thought this was (obviously) apparent when I said "analog data"
which differentiates it from "voice".

NOTE-  voice is not "analog data", per se, wherein data is defined
as digital computer data stream converted from Digital (inside the computer)
to analog at a **high speed**.. one does not usually consider the
voice as "data" although in a purely Spockian sense it is included
in the (marginal) definition of data transmission set theory.
A modem MOdulates and DEModulates these signals. Thus
we are speaking of two different things here- data rate as well
as data content. Voice does not meet the criteria I stated as to
either content or data rate, clearly.

Heck- u can use two Pepsi (coke, beans, ..) cans (from the left over
stove projects!! hah hah ha) and some string to transmit voice..
but u can't do computers over a string!!!)

**SOME** of the smaller companies competing with the Ma Bells
use an inexpensive CSS (central switching system) wherein
the A/D (analog to digital conversion circuitry and chips)
in their new digital phone lines (acoustic OR direct connection) just dont
switch fast enuf to capture the FSK (Frequency shift keying)- which is old
modem
technology BTW- newer modem technology uses stuff (similar
to spread spectrum ) like X2 standard--that is why we can get 56Kbs
on twisted pair- even 128KBS on ordinary copper in DSL..which
is  even much more demanding on the poor A/D's. See the patents
for the X2 and KFLEX and greater specs for details.

Also- Some of these use *2* (TWO!!) A/D's which completely hoses 56K and
u can only use V.34 (33.6k max), if that!!!  1200 again may be needed. Some
out of the way payphones work similar to the PBX's in your office and
are prevelant in smaller (read that cheaper) phone systems.

Another problem is that prior to February, 1998, there was no signle set
standard for 56K modems. You had a choice between two
proprietary 56K protocols: X2 from U.S. Robotics or K56flex
from Rockwell  (and Lucent I believe- I dont recall preceisely).

This could also lead to difficulties with the Sharps and others depending
on the specific specifications of your ISP and your internal (or external)
modem.

These items most likely constitute your problem with the Sharp.

The only 100% guaranteed reliable test is to get a 56K modem, get your
Sharp specs,
 and dial into your local ISP from a direct line if possible from your home
or office.
Be sure to install the drivers that come with the modem. Test each modem with
your ISP and if it works in your home base area and not in the "outback" you
know u have hit on the problem.

If it is your ISP u can yell at them-- a lot of places will carry both
Specs today.
Rockwell has had a new dual chip set out for sometime that will handle both
specs and even allows for future expansion (megabits of on-board flash memory).

Oh- a side note: Some of these cheaper switches are making it into the
mainstream
(ie Candella for example).  However, Ma Bell's (and others') digital lines
should
work fine as long as they only have 1 A/D between u and their outgoing lines.

Dont forget, a lot of these cheapoes switches now a days  built by startups
barely have any (audio or otherwise) bandwidth. Needless to
say they are not surviving.  One board I tested could handle 56kbs
direct but NOT 28kbs acoustic.

I suspect if one were still able to communicate at 300 baud or even 1200
it would work on the dig lines. If I ever get back into that field again
I can test a few boards I have laying around.

Last but not least, the audio couple is pretty much the last thing to
suspect (sorry)
as  the amplitude of all the units we tested were more than adequate to
drive the phone modulator, even with the new rectangular shaped phone
pieces (but I have not seen them out in the boonies yet <g> ).

--Rich

PS-

Least:

Above a certain threshold (called Shannon's Limit) the signal-to-noise
ratio of any medium becomes too low to reliably transfer data.

(such as the PCT flames- when Brick yells at us. That is noise.--oops
not his yelling I mean. He works to lessen the S/N ratio!).

The analog phone line and the Analog-to-Digital convertors are
the limiting factor in the speed of data (digital and analog)
transmission because of the  noise on the line (static, bad connections,
corroded wires under  screws and in slip-connectors after many years in
the elements).


PPS-

If you still r not convinced I suggest we take this debate line. If I have
time I
can clarify any engineering points.




At 8:04 AM -0800 3/5/01, Brick Robbins wrote:
>At 02:48 AM 3/5/01 , R.J.Calliger wrote:
>>Any analog based modem device, not just the Sharp, will unfortunately not
>>work with digital phone lines as these new lines cannot "speak"
>>analog data which has a series of tones used to represent computer
>>data.
>
>Not true in this case.
>
>Your voice is analog data. The phone seems to work with it OK. <g>
>
>The analog/digital problem you reference comes with direct wire connection
>from the analog modem to the digital phone line.
>
>The TM20 uses an acoustic interface, not a direct wire connection, so the
>FSK audio signal it uses should work on digital phone lines just fine.
>However it is fairly slow (which is OK for email).
>
>I think the problems with some pay phones has to do with the shape of the
>handset an the type of microphone and speakers they use, not the base
>technology.
>
>-B
>
>_______________________________________________
>PCT-L mailing list
>PCT-L@mailman.backcountry.net
>http://mailman.backcountry.net/mailman/listinfo/pct-l