Lucent winmodem connection rate drop problem

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Tue, 07 Aug 2001 19:37:36



Perhaps this message should be splitted among smaller ones, but anyway
here it goes... Just moved to a new house. A 16-meter phone cord was
installed to connect my Zoom, Lucent-based winmodem (Lucent generic
drivers ver. 6.00 installed) to the wall
jack. A check was performed that the connection is OK. A 50666 bps
connection was established on the first try. Just to make sure, I
tried to connect a couple of more times. Same figures.

Then a high WAF* figure is raised and the phone cable has to follow an
invisible route, hiding behind furniture and the rest. Since the cable
can not reach my system now, 5 more meters are bought and hand-wired
to the the former's end. The other end of the new cable is plugged to
the line input of my modem. Furthermore, a couple of other changes
took place:

1) A small plastic channel was installed to carry the phone cable and a
shielded TV antenna (RF modulated) cable. The RF cable runs in
parallel with the modem one for approximately 5 meters.

2) A second phone (DECT technology) was connected to the wall jack.

To make a long story short, I tried to connect again. To my
dissappointment, connection figures of 48000 bps were reported this
time, while V.34 connections were also made from time to time. The
48000 bps rate does not change during the call, but the
transmit rate is changed 28800 to 31200, per information provided by
ATI11. Additionally, ATI11 "Estimated Noise Level" is usually in the
280 region (wish someone could provide some info on this figure for the
specific rate: good? bad?). No RBS/digital pads present.

Backtracking into what has changed I unplug the antenna
cable from both ends, as well as the DECT telephone from the wall jack
(however, it is left "active", i.e. the DECT base is operating). What
else could have affected the connect rate? Would any post-call
diagnostics help identify the cuplrit?

PS: WAF: Wife Acceptance Factor
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Hooda Ges » Tue, 07 Aug 2001 21:30:43


From what you've written, I'd say it is the additional 5 meter extension.
Not sure what you meant by "hand-wired" unless you are using Cat3 or Cat5
cable and using heatshrink couplers (or twisting wires together with a dab
of solder). I'd suggest a single piece of Cat3 or 5 cable the full length
you will need and run it along the baseboard as best as you can, with *no*
other cables riding alongside of it. But 48000 is not a bad connect speed on
its own and more emphasis should be given to throughput than to connect
speed. If the modem is performing well and not getting dragged down by
errors and the wife is mollified by the current arrangement, I'd say count
your blessings. :)

Hooda


Quote:> Perhaps this message should be splitted among smaller ones, but anyway
> here it goes... Just moved to a new house. A 16-meter phone cord was
> installed to connect my Zoom, Lucent-based winmodem (Lucent generic
> drivers ver. 6.00 installed) to the wall
> jack. A check was performed that the connection is OK. A 50666 bps
> connection was established on the first try. Just to make sure, I
> tried to connect a couple of more times. Same figures.

> Then a high WAF* figure is raised and the phone cable has to follow an
> invisible route, hiding behind furniture and the rest. Since the cable
> can not reach my system now, 5 more meters are bought and hand-wired
> to the the former's end. The other end of the new cable is plugged to
> the line input of my modem. Furthermore, a couple of other changes
> took place:

> 1) A small plastic channel was installed to carry the phone cable and a
> shielded TV antenna (RF modulated) cable. The RF cable runs in
> parallel with the modem one for approximately 5 meters.

> 2) A second phone (DECT technology) was connected to the wall jack.

> To make a long story short, I tried to connect again. To my
> dissappointment, connection figures of 48000 bps were reported this
> time, while V.34 connections were also made from time to time. The
> 48000 bps rate does not change during the call, but the
> transmit rate is changed 28800 to 31200, per information provided by
> ATI11. Additionally, ATI11 "Estimated Noise Level" is usually in the
> 280 region (wish someone could provide some info on this figure for the
> specific rate: good? bad?). No RBS/digital pads present.

> Backtracking into what has changed I unplug the antenna
> cable from both ends, as well as the DECT telephone from the wall jack
> (however, it is left "active", i.e. the DECT base is operating). What
> else could have affected the connect rate? Would any post-call
> diagnostics help identify the cuplrit?

> PS: WAF: Wife Acceptance Factor
> --
> Michael.-
> -------------------------------------------------------------
> Remove "no" and "spam" from mail address for an e-mail reply.


 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Tue, 07 Aug 2001 22:29:34


| From what you've written, I'd say it is the additional 5 meter extension.
| Not sure what you meant by "hand-wired" unless you are using Cat3 or Cat5
| cable and using heatshrink couplers (or twisting wires together with a dab
| of solder). I'd suggest a single piece of Cat3 or 5 cable the full length
| you will need and run it along the baseboard as best as you can, with *no*
| other cables riding alongside of it. But 48000 is not a bad connect speed on
| its own and more emphasis should be given to throughput than to connect
| speed. If the modem is performing well and not getting dragged down by
| errors and the wife is mollified by the current arrangement, I'd say count
| your blessings. :)
|
| Hooda
|

Sorry for the confusing term used: by "hand-wired" I meant that I
stripped one end from each cable. Here, a phone cable carries 2 pairs
of conductors. For a single phone line only one pair is used. So I
connected the useful pair from the original cable to the corresponding
one from the extension, by twisting each conductor from the pair to
its corresponding one and putting some adhesive tape to keep them
together. Low-tech solution, which I also believe is the actual
problem. I wanted to get a single cable, but that would mean that I'd
have to rip the entire setup apart. Oh well, some things are simply
unavoidable :(

With regard to the throughput: if the cause of interference is
eliminated, then I'd expect that the rest will follow. Either a higher
connection rate, or the same one, albeit with a smaller estimated
noise level, which could yield smaller error figures on a physical
(electrical) level and, consequently, on a LAPM level.

I'll try cabling tomorrow and let you know of the results. Thanks!
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Hooda Ges » Wed, 08 Aug 2001 01:34:39




> | From what you've written, I'd say it is the additional 5 meter
extension.
> | Not sure what you meant by "hand-wired" unless you are using Cat3 or
Cat5
> | cable and using heatshrink couplers (or twisting wires together with a
dab
> | of solder). I'd suggest a single piece of Cat3 or 5 cable the full
length
> | you will need and run it along the baseboard as best as you can, with
*no*
> | other cables riding alongside of it. But 48000 is not a bad connect
speed on
> | its own and more emphasis should be given to throughput than to connect
> | speed. If the modem is performing well and not getting dragged down by
> | errors and the wife is mollified by the current arrangement, I'd say
count
> | your blessings. :)
> |
> | Hooda
> |

> Sorry for the confusing term used: by "hand-wired" I meant that I
> stripped one end from each cable. Here, a phone cable carries 2 pairs
> of conductors. For a single phone line only one pair is used. So I
> connected the useful pair from the original cable to the corresponding
> one from the extension, by twisting each conductor from the pair to
> its corresponding one and putting some adhesive tape to keep them
> together. Low-tech solution, which I also believe is the actual
> problem. I wanted to get a single cable, but that would mean that I'd
> have to rip the entire setup apart. Oh well, some things are simply
> unavoidable :(

It might have been a bit better to pick up some heatshrink splices and use
those. These are little plastic tubes (optionally with solder imbedded), for
those who are unfamiliar with them. They can be found at Radio Shack. Would
they make a better connection for data purposes? I don't know and suspect
not but they won't hurt and they will be more secure.

I'd have looked for a way to run new wiring to a closer location to the
computer. Cat3 cable run through the attic, for example, and dropped down
through the wall next to the computer. Bribing a small child to battle
snakes and crawl under the house dragging the wire to the location you want,
etc...

I've also run cabling along the baseboard, through walls, and so forth to
get a standard jack located close to my computers. I don't like to use teh
extension cables unless I can confirm they are high quality twisted pair
types.

Quote:

> With regard to the throughput: if the cause of interference is
> eliminated, then I'd expect that the rest will follow. Either a higher
> connection rate, or the same one, albeit with a smaller estimated
> noise level, which could yield smaller error figures on a physical
> (electrical) level and, consequently, on a LAPM level.

> I'll try cabling tomorrow and let you know of the results. Thanks!

Compare the throughput figures of before and after along with error rates
(in post-call diagnostics) and ignore connect speed.

Hooda

 
 
 

Lucent winmodem connection rate drop problem

Post by Art Jackso » Wed, 08 Aug 2001 02:27:44






> > | From what you've written, I'd say it is the additional 5 meter
> extension.
> > | Not sure what you meant by "hand-wired" unless you are using Cat3 or
> Cat5
> > | cable and using heatshrink couplers (or twisting wires together with a
> dab
> > | of solder). I'd suggest a single piece of Cat3 or 5 cable the full
> length
> > | you will need and run it along the baseboard as best as you can, with
> *no*
> > | other cables riding alongside of it. But 48000 is not a bad connect
> speed on
> > | its own and more emphasis should be given to throughput than to connect
> > | speed. If the modem is performing well and not getting dragged down by
> > | errors and the wife is mollified by the current arrangement, I'd say
> count
> > | your blessings. :)
> > |
> > | Hooda
> > |

> > Sorry for the confusing term used: by "hand-wired" I meant that I
> > stripped one end from each cable. Here, a phone cable carries 2 pairs
> > of conductors. For a single phone line only one pair is used. So I
> > connected the useful pair from the original cable to the corresponding
> > one from the extension, by twisting each conductor from the pair to
> > its corresponding one and putting some adhesive tape to keep them
> > together. Low-tech solution, which I also believe is the actual
> > problem. I wanted to get a single cable, but that would mean that I'd
> > have to rip the entire setup apart. Oh well, some things are simply
> > unavoidable :(

> It might have been a bit better to pick up some heatshrink splices and use
> those. These are little plastic tubes (optionally with solder imbedded), for
> those who are unfamiliar with them. They can be found at Radio Shack. Would
> they make a better connection for data purposes? I don't know and suspect
> not but they won't hurt and they will be more secure.

> I'd have looked for a way to run new wiring to a closer location to the
> computer. Cat3 cable run through the attic, for example, and dropped down
> through the wall next to the computer. Bribing a small child to battle
> snakes and crawl under the house dragging the wire to the location you want,
> etc...

> I've also run cabling along the baseboard, through walls, and so forth to
> get a standard jack located close to my computers. I don't like to use teh
> extension cables unless I can confirm they are high quality twisted pair
> types.

> > With regard to the throughput: if the cause of interference is
> > eliminated, then I'd expect that the rest will follow. Either a higher
> > connection rate, or the same one, albeit with a smaller estimated
> > noise level, which could yield smaller error figures on a physical
> > (electrical) level and, consequently, on a LAPM level.

> > I'll try cabling tomorrow and let you know of the results. Thanks!

> Compare the throughput figures of before and after along with error rates
> (in post-call diagnostics) and ignore connect speed.

> Hooda

I don't know if there is a Radio Shack in Greece or not, but I do
agree that his problem is the wiring used. If possible the whole
run should be replaced with twisted pair wiring, CAT3 or CAT5 from
the present phone jack all the way to the modem's jack if possible.
Another alternative would be to get a WAF Waiver and move the computer
closer to the phone line's wall jack. Good luck

--
Art Jackson W4TOY Owensboro, KY USA
Mud Chapel Quartet, Southern Gospel Music  
When all else fails, read THE instructions.
Found in the HOLY BIBLE       God Bless
Tired of all the changes? Read Hebrews 13:8

 
 
 

Lucent winmodem connection rate drop problem

Post by Hooda Ges » Wed, 08 Aug 2001 04:03:50







> > > | From what you've written, I'd say it is the additional 5 meter
> > extension.
> > > | Not sure what you meant by "hand-wired" unless you are using Cat3 or
> > Cat5
> > > | cable and using heatshrink couplers (or twisting wires together with
a
> > dab
> > > | of solder). I'd suggest a single piece of Cat3 or 5 cable the full
> > length
> > > | you will need and run it along the baseboard as best as you can,
with
> > *no*
> > > | other cables riding alongside of it. But 48000 is not a bad connect
> > speed on
> > > | its own and more emphasis should be given to throughput than to
connect
> > > | speed. If the modem is performing well and not getting dragged down
by
> > > | errors and the wife is mollified by the current arrangement, I'd say
> > count
> > > | your blessings. :)
> > > |
> > > | Hooda
> > > |

> > > Sorry for the confusing term used: by "hand-wired" I meant that I
> > > stripped one end from each cable. Here, a phone cable carries 2 pairs
> > > of conductors. For a single phone line only one pair is used. So I
> > > connected the useful pair from the original cable to the corresponding
> > > one from the extension, by twisting each conductor from the pair to
> > > its corresponding one and putting some adhesive tape to keep them
> > > together. Low-tech solution, which I also believe is the actual
> > > problem. I wanted to get a single cable, but that would mean that I'd
> > > have to rip the entire setup apart. Oh well, some things are simply
> > > unavoidable :(

> > It might have been a bit better to pick up some heatshrink splices and
use
> > those. These are little plastic tubes (optionally with solder imbedded),
for
> > those who are unfamiliar with them. They can be found at Radio Shack.
Would
> > they make a better connection for data purposes? I don't know and
suspect
> > not but they won't hurt and they will be more secure.

> > I'd have looked for a way to run new wiring to a closer location to the
> > computer. Cat3 cable run through the attic, for example, and dropped
down
> > through the wall next to the computer. Bribing a small child to battle
> > snakes and crawl under the house dragging the wire to the location you
want,
> > etc...

> > I've also run cabling along the baseboard, through walls, and so forth
to
> > get a standard jack located close to my computers. I don't like to use
teh
> > extension cables unless I can confirm they are high quality twisted pair
> > types.
[trimmed]

> I don't know if there is a Radio Shack in Greece or not, but I do
> agree that his problem is the wiring used. If possible the whole
> run should be replaced with twisted pair wiring, CAT3 or CAT5 from
> the present phone jack all the way to the modem's jack if possible.

Sure, the Radio Shack will be right near the McDonald's :) but point taken,
though any decent electrical/electronics supply shop will have these
devices.

Yes, replacing the entire cable would be best and running it anyway you can,
avoiding RF sources and electrical fields along the way.

Quote:> Another alternative would be to get a WAF Waiver and move the computer
> closer to the phone line's wall jack. Good luck

A WAF Waiver? Surely you jest...

Just remember, to have a successful marriage requires only two words from
the male... "yes, dear"

Hooda

 
 
 

Lucent winmodem connection rate drop problem

Post by Art Jackso » Wed, 08 Aug 2001 04:50:02









> > > > | From what you've written, I'd say it is the additional 5 meter
> > > extension.
> > > > | Not sure what you meant by "hand-wired" unless you are using Cat3 or
> > > Cat5
> > > > | cable and using heatshrink couplers (or twisting wires together with
> a
> > > dab
> > > > | of solder). I'd suggest a single piece of Cat3 or 5 cable the full
> > > length
> > > > | you will need and run it along the baseboard as best as you can,
> with
> > > *no*
> > > > | other cables riding alongside of it. But 48000 is not a bad connect
> > > speed on
> > > > | its own and more emphasis should be given to throughput than to
> connect
> > > > | speed. If the modem is performing well and not getting dragged down
> by
> > > > | errors and the wife is mollified by the current arrangement, I'd say
> > > count
> > > > | your blessings. :)
> > > > |
> > > > | Hooda
> > > > |

> > > > Sorry for the confusing term used: by "hand-wired" I meant that I
> > > > stripped one end from each cable. Here, a phone cable carries 2 pairs
> > > > of conductors. For a single phone line only one pair is used. So I
> > > > connected the useful pair from the original cable to the corresponding
> > > > one from the extension, by twisting each conductor from the pair to
> > > > its corresponding one and putting some adhesive tape to keep them
> > > > together. Low-tech solution, which I also believe is the actual
> > > > problem. I wanted to get a single cable, but that would mean that I'd
> > > > have to rip the entire setup apart. Oh well, some things are simply
> > > > unavoidable :(

> > > It might have been a bit better to pick up some heatshrink splices and
> use
> > > those. These are little plastic tubes (optionally with solder imbedded),
> for
> > > those who are unfamiliar with them. They can be found at Radio Shack.
> Would
> > > they make a better connection for data purposes? I don't know and
> suspect
> > > not but they won't hurt and they will be more secure.

> > > I'd have looked for a way to run new wiring to a closer location to the
> > > computer. Cat3 cable run through the attic, for example, and dropped
> down
> > > through the wall next to the computer. Bribing a small child to battle
> > > snakes and crawl under the house dragging the wire to the location you
> want,
> > > etc...

> > > I've also run cabling along the baseboard, through walls, and so forth
> to
> > > get a standard jack located close to my computers. I don't like to use
> teh
> > > extension cables unless I can confirm they are high quality twisted pair
> > > types.
> [trimmed]

> > I don't know if there is a Radio Shack in Greece or not, but I do
> > agree that his problem is the wiring used. If possible the whole
> > run should be replaced with twisted pair wiring, CAT3 or CAT5 from
> > the present phone jack all the way to the modem's jack if possible.

> Sure, the Radio Shack will be right near the McDonald's :) but point taken,
> though any decent electrical/electronics supply shop will have these
> devices.

> Yes, replacing the entire cable would be best and running it anyway you can,
> avoiding RF sources and electrical fields along the way.

> > Another alternative would be to get a WAF Waiver and move the computer
> > closer to the phone line's wall jack. Good luck

> A WAF Waiver? Surely you jest...

> Just remember, to have a successful marriage requires only two words from
> the male... "yes, dear"

> Hooda

Uh huh, I hear you. BTW, any problem with all the rain there?

--
Art Jackson W4TOY Owensboro, KY USA
Mud Chapel Quartet, Southern Gospel Music  
When all else fails, read THE instructions.
Found in the HOLY BIBLE       God Bless
Tired of all the changes? Read Hebrews 13:8

 
 
 

Lucent winmodem connection rate drop problem

Post by Hooda Ges » Wed, 08 Aug 2001 10:25:31



[trimmed]

Quote:> > > Another alternative would be to get a WAF Waiver and move the computer
> > > closer to the phone line's wall jack. Good luck

> > A WAF Waiver? Surely you jest...

> > Just remember, to have a successful marriage requires only two words
from
> > the male... "yes, dear"

> > Hooda

> Uh huh, I hear you. BTW, any problem with all the rain there?

No problems at all, surprisingly.  Three years ago, I had a moat but this
time nothing bad at all, no trouble  on the roads to work, no flooding of my
unpaved access road, my pond is still 50 feet from the house, and the gators
haven't eaten the cat.

Hooda

 
 
 

Lucent winmodem connection rate drop problem

Post by Alan Fowl » Wed, 08 Aug 2001 16:03:21



>Perhaps this message should be splitted among smaller ones, but anyway
>here it goes... Just moved to a new house. A 16-meter phone cord was
>installed to connect my Zoom, Lucent-based winmodem (Lucent generic
>drivers ver. 6.00 installed) to the wall
>jack. A check was performed that the connection is OK. A 50666 bps
>connection was established on the first try. Just to make sure, I
>tried to connect a couple of more times. Same figures.

>Then a high WAF* figure is raised and the phone cable has to follow an
>invisible route, hiding behind furniture and the rest. Since the cable
>can not reach my system now, 5 more meters are bought and hand-wired
>to the the former's end. The other end of the new cable is plugged to
>the line input of my modem. Furthermore, a couple of other changes
>took place:

>1) A small plastic channel was installed to carry the phone cable and a
>shielded TV antenna (RF modulated) cable. The RF cable runs in
>parallel with the modem one for approximately 5 meters.

>2) A second phone (DECT technology) was connected to the wall jack.

Q'  What is in the DECT base?  Could it be generating any
interfering signals?  Is it connected directly to the phone
line or plugged into the phone jack?  What is DECT
technology please?  I don't recognise the term.

Q.  Is there any power wiring in the wall behind the new
cable run?

        I don't like to see "floating wires in a cable.
They should be earthed if possible (without interfering with
the telephone wiring)

- Show quoted text -

Quote:

>To make a long story short, I tried to connect again. To my
>dissappointment, connection figures of 48000 bps were reported this
>time, while V.34 connections were also made from time to time. The
>48000 bps rate does not change during the call, but the
>transmit rate is changed 28800 to 31200, per information provided by
>ATI11. Additionally, ATI11 "Estimated Noise Level" is usually in the
>280 region (wish someone could provide some info on this figure for the
>specific rate: good? bad?). No RBS/digital pads present.

>Backtracking into what has changed I unplug the antenna
>cable from both ends, as well as the DECT telephone from the wall jack
>(however, it is left "active", i.e. the DECT base is operating). What
>else could have affected the connect rate? Would any post-call
>diagnostics help identify the cuplrit?

>PS: WAF: Wife Acceptance Factor

 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Wed, 08 Aug 2001 16:50:12





| >
| > | From what you've written, I'd say it is the additional 5 meter
| extension.
| > | Not sure what you meant by "hand-wired" unless you are using Cat3 or
| Cat5
| > | cable and using heatshrink couplers (or twisting wires together with a
| dab
| > | of solder). I'd suggest a single piece of Cat3 or 5 cable the full
| length
| > | you will need and run it along the baseboard as best as you can, with
| *no*
| > | other cables riding alongside of it. But 48000 is not a bad connect
| speed on
| > | its own and more emphasis should be given to throughput than to connect
| > | speed. If the modem is performing well and not getting dragged down by
| > | errors and the wife is mollified by the current arrangement, I'd say
| count
| > | your blessings. :)
| > |
| > | Hooda
| > |
| >
| > Sorry for the confusing term used: by "hand-wired" I meant that I
| > stripped one end from each cable. Here, a phone cable carries 2 pairs
| > of conductors. For a single phone line only one pair is used. So I
| > connected the useful pair from the original cable to the corresponding
| > one from the extension, by twisting each conductor from the pair to
| > its corresponding one and putting some adhesive tape to keep them
| > together. Low-tech solution, which I also believe is the actual
| > problem. I wanted to get a single cable, but that would mean that I'd
| > have to rip the entire setup apart. Oh well, some things are simply
| > unavoidable :(
|
| It might have been a bit better to pick up some heatshrink splices and use
| those. These are little plastic tubes (optionally with solder imbedded), for
| those who are unfamiliar with them. They can be found at Radio Shack. Would
| they make a better connection for data purposes? I don't know and suspect
| not but they won't hurt and they will be more secure.
|
| I'd have looked for a way to run new wiring to a closer location to the
| computer. Cat3 cable run through the attic, for example, and dropped down
| through the wall next to the computer. Bribing a small child to battle
| snakes and crawl under the house dragging the wire to the location you want,
| etc...

No problem, my daughther could surely do that. Quite inexpensive too
(only some candies required)! If only I had an attic... This is an
apartment, so I do not have plenty of choices.

|
| I've also run cabling along the baseboard, through walls, and so forth to
| get a standard jack located close to my computers. I don't like to use teh
| extension cables unless I can confirm they are high quality twisted pair
| types.
|
|
| >
| > With regard to the throughput: if the cause of interference is
| > eliminated, then I'd expect that the rest will follow. Either a higher
| > connection rate, or the same one, albeit with a smaller estimated
| > noise level, which could yield smaller error figures on a physical
| > (electrical) level and, consequently, on a LAPM level.
| >
| > I'll try cabling tomorrow and let you know of the results. Thanks!
|
|
| Compare the throughput figures of before and after along with error rates
| (in post-call diagnostics) and ignore connect speed.
|

Took a single 25 meter cable and tested it on a straight route from
wall jack to modem. Same connection rate _and_ throughput. Estimated
noise level dropped slightly to 265-272. Vast improvement though on
the number of v.34 connections established (0) over a number of
dial-up attempts. V.90/48000 bps straight each time, with minimal
negotiation after the DIL is sent by the digital modem. Quite
encouraging.

Yet, I did make the mistake of taking plain phone cable. I know that
CAT3/CAT5/twisted pair are used in LANs, but which one provides best
shielding/noise attenutation characteristics? Furthermore, don't these
cables use 3-4 pairs internally? Here the phone jacks use RJ-11
connectors, so how would one apply a (two-pair) RJ11 connector on a
cable carrying more than two pairs of conductors?
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Wed, 08 Aug 2001 17:15:22





| > >


[...]

| > Another alternative would be to get a WAF Waiver and move the computer
| > closer to the phone line's wall jack. Good luck

Good luck? Luck is not an issue here, believe me :) Picture someone
with * springing out of his ears to get the idea of confronting a
furniture-loving, cabling-loathing woman, no matter what my arguments
are. Alas, I just can't have it all my way.

|
| A WAF Waiver? Surely you jest...
|
| Just remember, to have a successful marriage requires only two words from
| the male... "yes, dear"

Three words _at least_ in Greek actually, which brings Greece and the
States one more step apart. Throw in the lack of Radio Shack stores
here and we are talking about severely different civilizations ;)
Remember also that Greeks are Mediterranean cultured to the bone, so,
like our friends the Italians, three words just don't cut it. Throw
three pages of affirmative responses to get a better idea of
male-female interactions here.

Note that I am just joking here (my wife does not read NGs, but then
again, a "better safe than sorry" attitude seems like a good backup).
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Wed, 08 Aug 2001 17:31:32



|
| >Perhaps this message should be splitted among smaller ones, but anyway
| >here it goes... Just moved to a new house. A 16-meter phone cord was
| >installed to connect my Zoom, Lucent-based winmodem (Lucent generic
| >drivers ver. 6.00 installed) to the wall
| >jack. A check was performed that the connection is OK. A 50666 bps
| >connection was established on the first try. Just to make sure, I
| >tried to connect a couple of more times. Same figures.
| >
| >Then a high WAF* figure is raised and the phone cable has to follow an
| >invisible route, hiding behind furniture and the rest. Since the cable
| >can not reach my system now, 5 more meters are bought and hand-wired
| >to the the former's end. The other end of the new cable is plugged to
| >the line input of my modem. Furthermore, a couple of other changes
| >took place:
| >
| >1) A small plastic channel was installed to carry the phone cable and a
| >shielded TV antenna (RF modulated) cable. The RF cable runs in
| >parallel with the modem one for approximately 5 meters.
| >
| >2) A second phone (DECT technology) was connected to the wall jack.
|
| Q'  What is in the DECT base?  Could it be generating any
| interfering signals?  Is it connected directly to the phone
| line or plugged into the phone jack?  What is DECT
| technology please?  I don't recognise the term.
|

DECT stands for Digital Enhanced Cordless Telephony, if I recall
correctly. It uses some bands in the 1800Mhz region to transmit voice
using digital techniques over ranges up to 300 meteres. DECT stations
(bases) can also interwork in a limited way, like mobile radio
stations do. When off-hook, audible interference is generated on other
phone equipment, but when on-hook communucation between the base and
the headset is limited to occasional traffic (like obtaining the
headset's distance to tune signal strength). I'll unplug the base,
switch off the headset and see what happens.

| Q.  Is there any power wiring in the wall behind the new
| cable run?
|

I installed a single phone cable, that runs straight to the system
avoiding any (visible, at least) power source. No change in connection
rate. But no V.34 connections either, which is certainly a step in the
right direction.

|       I don't like to see "floating wires in a cable.
| They should be earthed if possible (without interfering with
| the telephone wiring)

You are right. Perhaps I'll also replace this cable by a CAT3/CAT5 one.

| >
| >To make a long story short, I tried to connect again. To my
| >dissappointment, connection figures of 48000 bps were reported this
| >time, while V.34 connections were also made from time to time. The
| >48000 bps rate does not change during the call, but the
| >transmit rate is changed 28800 to 31200, per information provided by
| >ATI11. Additionally, ATI11 "Estimated Noise Level" is usually in the
| >280 region (wish someone could provide some info on this figure for the
| >specific rate: good? bad?). No RBS/digital pads present.
| >
| >Backtracking into what has changed I unplug the antenna
| >cable from both ends, as well as the DECT telephone from the wall jack
| >(however, it is left "active", i.e. the DECT base is operating). What
| >else could have affected the connect rate? Would any post-call
| >diagnostics help identify the cuplrit?
| >
| >PS: WAF: Wife Acceptance Factor
|

--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Hooda Ges » Wed, 08 Aug 2001 21:18:40




[trimmed]
> |
> | I'd have looked for a way to run new wiring to a closer location to the
> | computer. Cat3 cable run through the attic, for example, and dropped
down
> | through the wall next to the computer. Bribing a small child to battle
> | snakes and crawl under the house dragging the wire to the location you
want,
> | etc...

> No problem, my daughther could surely do that. Quite inexpensive too
> (only some candies required)! If only I had an attic... This is an
> apartment, so I do not have plenty of choices.

Apartments are a bit rougher. The owners tend to frown on drilling through
walls to run phone cables (but I have done it, just make it as small as
possible and where it won't show). In the US, however, most apartments are
wired for both cable TV and phone in each bedroom so it isn't as much of an
issue anymore.

Quote:> |
> | Compare the throughput figures of before and after along with error
rates
> | (in post-call diagnostics) and ignore connect speed.
> |

> Took a single 25 meter cable and tested it on a straight route from
> wall jack to modem. Same connection rate _and_ throughput. Estimated
> noise level dropped slightly to 265-272. Vast improvement though on
> the number of v.34 connections established (0) over a number of
> dial-up attempts. V.90/48000 bps straight each time, with minimal
> negotiation after the DIL is sent by the digital modem. Quite
> encouraging.

Yes, that fairly confirms the extra 5 meter length (or the coupling method)
caused the most problems.

Quote:

> Yet, I did make the mistake of taking plain phone cable. I know that
> CAT3/CAT5/twisted pair are used in LANs, but which one provides best
> shielding/noise attenutation characteristics? Furthermore, don't these
> cables use 3-4 pairs internally? Here the phone jacks use RJ-11
> connectors, so how would one apply a (two-pair) RJ11 connector on a
> cable carrying more than two pairs of conductors?

The Cat3 comes in both 2 and 4 pair forms, I believe. The Cat5 is usually
found with 4 pair. Either type will be more than adequate. Just separate the
pair(s) needed and attach the connectors. I have, sitting in a drawer, a 30
foot (9 meters) length of Cat5 cable with RJ-11 connectors on each end
utilising the 2 center pair. It's not hard to do.

Hooda

 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Wed, 08 Aug 2001 22:26:12




[...]

| >
| > Yet, I did make the mistake of taking plain phone cable. I know that
| > CAT3/CAT5/twisted pair are used in LANs, but which one provides best
| > shielding/noise attenutation characteristics? Furthermore, don't these
| > cables use 3-4 pairs internally? Here the phone jacks use RJ-11
| > connectors, so how would one apply a (two-pair) RJ11 connector on a
| > cable carrying more than two pairs of conductors?
|
| The Cat3 comes in both 2 and 4 pair forms, I believe. The Cat5 is usually
| found with 4 pair. Either type will be more than adequate. Just separate the
| pair(s) needed and attach the connectors. I have, sitting in a drawer, a 30
| foot (9 meters) length of Cat5 cable with RJ-11 connectors on each end
| utilising the 2 center pair. It's not hard to do.

Luckily for me, stores are open this evening. A 25-meter CAT5,
shielded cable is in order. Third time (lucky, perhaps?) I buy a cable
for the modem this week. I'll have to sneak this one into the house
though, otherwise the WAF factor will kick in (or me, for that matter)
in a major way :)
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.

 
 
 

Lucent winmodem connection rate drop problem

Post by Michail Pappa » Thu, 09 Aug 2001 17:12:33




[...]

| > Took a single 25 meter cable and tested it on a straight route from
| > wall jack to modem. Same connection rate _and_ throughput. Estimated
| > noise level dropped slightly to 265-272. Vast improvement though on
| > the number of v.34 connections established (0) over a number of
| > dial-up attempts. V.90/48000 bps straight each time, with minimal
| > negotiation after the DIL is sent by the digital modem. Quite
| > encouraging.
|
| Yes, that fairly confirms the extra 5 meter length (or the coupling method)
| caused the most problems.

Had the chance yesterday to perform some extensive tests with a single
ISP. I used the single (plain) phone cable. Additionally, unplugged
the DECT phone from both power and the phone line. Success! 7 out 7
times connected straight at 52000 bps. Post-call diagnostics indicated
some rate renegotiations to 31200 or 50666, so I limited the max
received rate to 50666 and tried again. No retrains/renegs whatsoever!

From the 50666 connections checked, the following observations were
made:

a) The estimated noise level for most good noise connections was
_under_ 200, in the 165-185 range. Taking into account that higher
figures are worse and that for the same line conditions a higher
connect rate will produce a higher estimated noise level, the
previously encountered value of 270 for 48000 bps connections is
_very_ high. Hence, a _major_ improvement in connection rate. The
actual throughput confirmed line quality: downloading self-extracting
compressed .exe files at rates higher than 5.5-6 kbytes/sec. Notice
that the 50666 rate remained constant in all cases. Finally, received
power level was -8dBm.

b) For the not-so-good 50666 connections, estimated noise level was
higher than 220. BTW, it would be nice to have a formula that
outputs the estimated noise level, to use it for mapping quality
values from a connect rate to another. Receive power level was
slightly worse too, at -9dBm.

c) The same connections of (b) above indicated that robbed-bit
signalling was present in some cases! I certainly did not expect this,
since RBS signalling is inherent in T1 links which do not contain
a timeslot for the purpose, but not in E1 links used here.

d) The percentage of erroneous packets received to the total
number of packets received, served as an additional quality metric. In
good connections it ranged from 0 to 0.5%, while the rest exhibited
figures in the range 1-3%.

Encouraged by these results, I proceeded to connect the DECT phone to the
other wall phone jack. 50666 connections were possible, although
yesterday I could not connect at rates higher 48000 bps! _But_ the
error rate climbed above 250. In a case or two figures as high as 350
were also encountered. Throughput fell in a major way. Local modem
initiated retrains were exhibited too. The heck with it, I thought,
and connected the DECT base on the phone jack of my modem. This phone
must generate a lot of interference since I was not able to connect
above  26400. Even at those rates actual throuput was 1-1.5 kbytes!

For the time I'll just replace the phone cable my modem uses, with a
shielded CAT5 one. I'll also use a 1-2 meter cable to connect the DECT
phone to the other phone jack. I still am afraid that the latter one
will be of no use since the DECT base produces RF noise at the source,
so CAT 5 shielding might not help.
--
Michael.-
-------------------------------------------------------------
Remove "no" and "spam" from mail address for an e-mail reply.