Modem question

On 09/07/2010 01:36 PM, Roman Rumian wrote:
> Hi Tim,
>
> W dniu 2010-09-07 22:01, Tim Wescott pisze:
> (...)
>>> what a maximum range is possible for 4800bps full duplex wire (a pair of
>>> 0.9 mmm wires) modem ?
>>
>> What standard?
>
> No standard, or anyone fitting the need.
>
>> How noisy is the environment?
>
> It's hard to say. We have 20+ kilometers of wire pair, and the customer
> wants to have duplex transmission at minimum 4800bps.
> Thirst idea is to measure frequency characteristic of this pair.
> Probably it is also possible to drive that pair with high amplitude
> signals (24 volts or more) to have better SNR.
>
>> For a custom made modem, with a comm format tailored to be whatever it
>> takes, just name your distance and start counting your money.
>>
>> Specific modem standards (i.e. ITU-T V.29) will have specific
>> characteristics, but even those will have performance that varies wildly
>> with the quality of the installation of the wire. If you use one of the
>> telephone specifications and you put it on telephone-grade wire with a
>> telephone-grade installation, you should be able to make it work over
>> many miles.
>
> How many ?
> Have found two pairs modem for 19200bps at distance 20+ miles (36km),
> but we have only one pair.
> I am searching for ready to use device/algorithm/solution, as I prefer
> other type of DSP development activity (professional audio). ;-)
>
>> Your best bet is to either look around the web, ask on a group where
>> folks actually use modems instead of trading information on how best to
>> design them, or attempt to buy a modem with a guarantee of how long a
>> wire it'll work on.
>
> Agree.
>
> Many thanks for your answer Tim. :-)

I've been assuming this is a one-off -- am I correct?  I'm trying to 
think of ways that you can do this with an equipment purchase, which 
should run to one or two zeros less in the price tag than a totally 
custom solution.

But I'm frustrated because while I know this is a solved problem in 
industrial circles I don't have the experience with it myself.  Give me 
your 20km of line and I'll design you a modem, sure -- but this has 
already been done!

You might try asking on sci.engr.control -- yours isn't a controls 
question, but if there are any industrial controls engineers left over 
there they may have worked on something like that.

sci.electronics.design may also yield some fruit, if you don't mind the 
US-specific political noise there.

What you're looking for is leased-line modem.  You could do this with 
ordinary telephone modems, but you'd need a central office simulator to 
make a ring signal so the modem on the other end would pick up (ugh), 
and you may need to do some compensation for the 20km of line.  A leased 
line modem pair should take care of that for you, for more money and 
hassle than a regular modem on a regular phone line would, but 
(hopefully!) for less money and hassle than a full custom solution.

Does this do what you want?
http://www.blackbox.com/Store/Detail.aspx/Industrial-Modem-288/MD3400A

(they have more -- and this sort of thing isn't their main product line).

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

eric.jacobsen@ieee.org (Eric Jacobsen) writes:
> [...]
> To get full duplex, though, you need to find something that does TDD
> or FDD within the capabilities of the wire.

Couldn't you use cancellation techniques?
-- 
Randy Yates                      % "Watching all the days go by...    
Digital Signal Labs              %  Who are you and who am I?"
mailto://yates@ieee.org          % 'Mission (A World Record)', 
http://www.digitalsignallabs.com % *A New World Record*, ELO

Tim Wescott <tim@seemywebsite.com> wrote:
(snip)
 
> What you're looking for is leased-line modem.  You could do this with 
> ordinary telephone modems, but you'd need a central office simulator to 
> make a ring signal so the modem on the other end would pick up (ugh), 
> and you may need to do some compensation for the 20km of line.  A leased 
> line modem pair should take care of that for you, for more money and 
> hassle than a regular modem on a regular phone line would, but 
> (hopefully!) for less money and hassle than a full custom solution.

There used to be some that could be put into answer mode without
requiring a ring signal.  I believe some of the old Hayes modems.

While others used the same command codes, I don't know that all
implemented that feature.

-- glen

On 09/07/2010 04:37 PM, glen herrmannsfeldt wrote:
> Tim Wescott<tim@seemywebsite.com>  wrote:
> (snip)
>
>> What you're looking for is leased-line modem.  You could do this with
>> ordinary telephone modems, but you'd need a central office simulator to
>> make a ring signal so the modem on the other end would pick up (ugh),
>> and you may need to do some compensation for the 20km of line.  A leased
>> line modem pair should take care of that for you, for more money and
>> hassle than a regular modem on a regular phone line would, but
>> (hopefully!) for less money and hassle than a full custom solution.
>
> There used to be some that could be put into answer mode without
> requiring a ring signal.  I believe some of the old Hayes modems.
>
> While others used the same command codes, I don't know that all
> implemented that feature.

If that's the case, all you need for a cheezy CO simulator is a 24 volt 
source and a 600 ohm resistor -- 48V if you want to be authentic.

  o-----------------o------------------o
                    |
                   .-.
                   | | 600
                   '-'
                    |
                   ---
                 -------
                   ---
                 -------
                    |
  o-----------------o-------------------o

There's fancier ways -- but this works quick.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

On Tue, 07 Sep 2010 19:10:08 -0400, Randy Yates <yates@ieee.org>
wrote:

>eric.jacobsen@ieee.org (Eric Jacobsen) writes:
>> [...]
>> To get full duplex, though, you need to find something that does TDD
>> or FDD within the capabilities of the wire.
>
>Couldn't you use cancellation techniques?
>-- 
>Randy Yates                      % "Watching all the days go by...    
>Digital Signal Labs              %  Who are you and who am I?"
>mailto://yates@ieee.org          % 'Mission (A World Record)', 
>http://www.digitalsignallabs.com % *A New World Record*, ELO

One probably could do that, but if he's looking for existing equipment
his selections will probably be limited.   I don't know of any general
purpose hardware that does that.   A few folks do it for satellite
modems, where the spectrum is really expensive, but other than that it
seems pretty rare.



Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.abineau.com

Eric Jacobsen <eric.jacobsen@ieee.org> wrote:
> On Tue, 07 Sep 2010 19:10:08 -0400, Randy Yates <yates@ieee.org>
>>> or FDD within the capabilities of the wire.
(snip)

>>Couldn't you use cancellation techniques?
(snip)
 
> One probably could do that, but if he's looking for existing equipment
> his selections will probably be limited.   I don't know of any general
> purpose hardware that does that.   A few folks do it for satellite
> modems, where the spectrum is really expensive, but other than that it
> seems pretty rare.

I thought that all the high speed, 9600 bit/sec and higher,
modems did it.  

Gigabit ethernet, 1000baseT, does.

-- glen

On Wed, 8 Sep 2010 00:33:45 +0000 (UTC), glen herrmannsfeldt
<gah@ugcs.caltech.edu> wrote:

>Eric Jacobsen <eric.jacobsen@ieee.org> wrote:
>> On Tue, 07 Sep 2010 19:10:08 -0400, Randy Yates <yates@ieee.org>
>>>> or FDD within the capabilities of the wire.
>(snip)
>
>>>Couldn't you use cancellation techniques?
>(snip)
> 
>> One probably could do that, but if he's looking for existing equipment
>> his selections will probably be limited.   I don't know of any general
>> purpose hardware that does that.   A few folks do it for satellite
>> modems, where the spectrum is really expensive, but other than that it
>> seems pretty rare.
>
>I thought that all the high speed, 9600 bit/sec and higher,
>modems did it.  
>
>Gigabit ethernet, 1000baseT, does.
>
>-- glen

Are they cancelling the back channel?   I.e., simultaneous Tx and Rx
on the same channel with Tx cancellation at the Rx?    I think that's
what Randy was asking about.    I wasn't aware that any wired
standards did that, if they do.


Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.abineau.com

>On Wed, 8 Sep 2010 00:33:45 +0000 (UTC), glen herrmannsfeldt
><gah@ugcs.caltech.edu> wrote:
>
>>Eric Jacobsen <eric.jacobsen@ieee.org> wrote:
>>> On Tue, 07 Sep 2010 19:10:08 -0400, Randy Yates <yates@ieee.org>
>>>>> or FDD within the capabilities of the wire.
>>(snip)
>>
>>>>Couldn't you use cancellation techniques?
>>(snip)
>> 
>>> One probably could do that, but if he's looking for existing equipment
>>> his selections will probably be limited.   I don't know of any general
>>> purpose hardware that does that.   A few folks do it for satellite
>>> modems, where the spectrum is really expensive, but other than that it
>>> seems pretty rare.
>>
>>I thought that all the high speed, 9600 bit/sec and higher,
>>modems did it.  
>>
>>Gigabit ethernet, 1000baseT, does.
>>
>>-- glen
>
>Are they cancelling the back channel?   I.e., simultaneous Tx and Rx
>on the same channel with Tx cancellation at the Rx?    I think that's
>what Randy was asking about.    I wasn't aware that any wired
>standards did that, if they do.

How do you think you can get 56kbps in one direction and 33.6kbps in the
other direction (V.90), into a 64kbps G.711 telephone circuit, without
using every scrap of the available band in both directions? Modems
typically apply 4 short cancellers, adaptively spread in time, to hit the 2
hybrids in the local analogue section, and the 2 hybrids in the remote
analogue section. Obviously the big digital chunk in the middle of the
channel doesn't need cancellation. The dispersed echo from discontinuities
along the analogue pairs doesn't contribute much.

Every new duplex telephone line modem standard since the early 80s relies
on high quality echo cancellation. The FAX modems - V.17 and V.34HDX - are,
I think, the only modern non-duplex modem standards, and the only ones
which do not rely on echo cancellation.

Steve

steveu wrote:

> Modems
> typically apply 4 short cancellers, adaptively spread in time, to hit the 2
> hybrids in the local analogue section, and the 2 hybrids in the remote
> analogue section. Obviously the big digital chunk in the middle of the
> channel doesn't need cancellation. The dispersed echo from discontinuities
> along the analogue pairs doesn't contribute much.

IIRC V.32 used frequency shifting for the far side echo, as the 
frequency shift could be introduced by the legacy analog channel 
concentration equipment. Don't know if they support for this feature in 
the later standards.

> Every new duplex telephone line modem standard since the early 80s relies
> on high quality echo cancellation. The FAX modems - V.17 and V.34HDX - are,
> I think, the only modern non-duplex modem standards, and the only ones
> which do not rely on echo cancellation.

I doubt feasibility of high quality echo cancellation in the bandwidth 
of tens of MHz.


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

>IIRC V.32 used frequency shifting for the far side echo, as the 
>frequency shift could be introduced by the legacy analog channel 
>concentration equipment. Don't know if they support for this feature in 
>the later standards.
>
>Vladimir Vassilevsky

All the modem standards up to V.32bis require that you tolerate a 8-10Hz
shift between the sending frequency and the received one. This is to allow
for FDM equipment in the channel, which was common until quite recently. I
think V.34 leaves the issue loose, and V.90/V.92 are only for pure digital
transmission networks.

FDM frequency errors mean the components of the echo of your own signal
from the remote hybrids could be frequency shifted. The components of the
echo from the local hybrids is not. Voice echo cancellers generally assume
no frequency shift, but modem cancellers have to allow for it to have any
chance of working through an FDM link.

The 4 short section sparse cancellers, typical in modems, treat the 2 local
sections and the 2 remote sections separately. The local hybrids are
cancelled between the transmitted and received signals. The remote hybrids
are cancelled between the transmitted signal, and a de-rotated version of
the received signal. Thus, a part of the canceller training is to evaluate
the frequency shift. The specs provide formal training procedures as a
connection is brought up, which makes the initial evaluation of frequency
shift and echo practical. I think its impractical to train one of these
cancellers on the real signal. Its a nice wide band signal, but there is
100% double talk.

Of course, a sparse canceller cuts computation and coefficient noise as
well, so you would use them with or without the FDM issue.

Steve