Hi
I have done some testing with a U.S.Robotics56K modem at a speed of
50666/31200 in buffer mode with no error control, and if I only are in
data phase and I'm not sending or receiving any data I can see that
I'm getting some noise or bursts that I'm not expecting. I don't think
all of this is line noise but I'm not 100% sure. What I think is that
this is some signaling between the digital modem and my modem? How
does the modem behave when it's in data phase?

U.S. Robotics 56K FAX EXT Link Diagnostics...


Modulation V.90
Carrier Freq (Hz) None/1920
Symbol Rate 8000/3200
Trellis Code None/64S-4D
Nonlinear Encoding None/ON
Precoding None/ON
Shaping ON/ON
Preemphasis (-dB) 0/2
Recv/Xmit Level (-dBm) 20/16
Near Echo Loss (dB) 18
Far Echo Loss (dB) 0
Carrier Offset (Hz) NONE
Round Trip Delay (msec) 89
Timing Offset (ppm) 3756
SNR (dB) 49.3
Speed Shifts Up/Down 0/0
Status : aa,5,13N,13.2,-0,0N,0,46.4,19.20

U.S. Robotics 56K FAX EXT Link Diagnostics...

Chars sent 0 Chars Received 22813
Chars lost 0
Octets sent 0 Octets Received 0
Blocks sent 0 Blocks Received 0
Blocks resent 0

Retrains Requested 0 Retrains Granted 0
Line Reversals 0 Blers 0
Link Timeouts 0 Link Naks 0

Data Compression NONE
Equalisation Long
Fallback Enabled
Protocol NONE
Speed 50666/31200
V.90 Peak Speed 50666
Last Call 00:20:36

Best regards,
Lars Homlung

Your connection looks dandy at the modulation layer,
but it seems that you have failed to negotiate error
control (LAPM or MNP4.) This will cause reduced performance,
and will cause errored data to be delivered to your application
unless you are using some error detecting/correcting protocol
at a higher layer (like PPP/IP/TCP.)

Indeed, if using a modern modulation like V.34 or V.90, it
is quite normal for a non error controlled modem link to deliver
errored data (even when "idle").

Can you please explain why you would not want to use error control??

Aaron

---

~ Hi
~ I have done some testing with a U.S.Robotics56K modem at a speed of
~ 50666/31200 in buffer mode with no error control, and if I only are in
~ data phase and I'm not sending or receiving any data I can see that
~ I'm getting some noise or bursts that I'm not expecting. I don't think
~ all of this is line noise but I'm not 100% sure. What I think is that
~ this is some signaling between the digital modem and my modem? How
~ does the modem behave when it's in data phase?
~
~ U.S. Robotics 56K FAX EXT Link Diagnostics...
~
~
~ Modulation V.90
~ Carrier Freq (Hz) None/1920
~ Symbol Rate 8000/3200
~ Trellis Code None/64S-4D
~ Nonlinear Encoding None/ON
~ Precoding None/ON
~ Shaping ON/ON
~ Preemphasis (-dB) 0/2
~ Recv/Xmit Level (-dBm) 20/16
~ Near Echo Loss (dB) 18
~ Far Echo Loss (dB) 0
~ Carrier Offset (Hz) NONE
~ Round Trip Delay (msec) 89
~ Timing Offset (ppm) 3756
~ SNR (dB) 49.3
~ Speed Shifts Up/Down 0/0
~ Status : aa,5,13N,13.2,-0,0N,0,46.4,19.20
~
~ U.S. Robotics 56K FAX EXT Link Diagnostics...
~
~ Chars sent 0 Chars Received 22813
~ Chars lost 0
~ Octets sent 0 Octets Received 0
~ Blocks sent 0 Blocks Received 0
~ Blocks resent 0
~
~ Retrains Requested 0 Retrains Granted 0
~ Line Reversals 0 Blers 0
~ Link Timeouts 0 Link Naks 0
~
~ Data Compression NONE
~ Equalisation Long
~ Fallback Enabled
~ Protocol NONE
~ Speed 50666/31200
~ V.90 Peak Speed 50666
~ Last Call 00:20:36
~
~ Best regards,
~ Lars Homlung

I did some modem tests on a network to a company that had some
problems when they used V.90 and V.34 modems. I did some changes to
the network, and to see how clean the connection was I run the modems
in buffermode and disabled the error/ flowcontrol. I got a clean
connection on the V.34 and on 40000 with the V.90 modem, but if I used
higher speeds on the V.90 I got some bursts (0.01% when I run traffic
for 20 min). I have tried to find out why this happens and I haven't
found a direct answer but I think it's because of the high sampling
rate and/or that the analog and the digital modem are doing some
speed/level negotiation.

Regards Lars

Aaron Leonard <Aaron@Cisco.COM> wrote in message news:<i6f1uvosjrk2rpbvi4479ohar4v1cnj4hl@4ax.com>. ..

lars.homlung@semcon.com (Lars Homlung) wrote:
The sampling rate is the same regardless of the speed. And almost
certainly it isn't due to speed/level negotiations.

The higher the speed the more unique values have to be recognized,
which means finer granularity in the PAM signal. That makes, at
the lower voltage steps, the noise on the cable much more significant
at higher speeds than at lower speeds. Another factor, which is
hard to estimate, might be the quality of the codec in the telco's
line card for any given line. The telco installs thousands of line
cards, and has no need to use high quality codecs that are intended
for low error rate data! Instead they save thousands of dollars by
using codecs that are good enough for voice grade... which means
that some of them are probably very good, and some aren't. Hence
it is possible that what you are seeing is also nothing more than
variations in the output voltage from the telco's codec.

Whatever, given a 50666 bps connection, I'm hard pressed to see
what point there is in wasting time figuring out why it isn't faster
or less "noisy". To put it mildly, you have a fabulous line.

--
Floyd L. Davidson <http://web.newsguy.com/floyd_davidson>
Ukpeagvik (Barrow, Alaska) floyd@barrow.com