Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[dizzy4528]

Hi guys , Bear with me as the question is a bit long winded.
Having set up with Newnet Max last week and originally connecting at 1900kbps
on a very long line, atten 63db. I was running at 6db of SNR till the SNR gradually fell to 0 and the router did a resync, this continued a number of times till I assume the level of 12 or15db kicked in, trying to stabilize my line.
In between this I tried a number of problem solving ideas recommended on this site including fitting a filter plate to the master socket, checking for the ring wire and connecting the router direct to the master socket with no further telephone extensions connected. This did not seem to cure the problem and eventually the sync speed dropped to around 500kbps.This is when my telephone line from the side of the house to the pole decided to give up the ghost and fail completely .This had been the problem for some time ,creating noise on the line which I admit I didn’t notice on my old fixed 512K speed service with a 23db noise margin .BT replaced the cable and low and behold the SNR remained at 11db never wavering from this by more than 1db,even through the night ,the resultant stabilisation allowed the sync speed to increase to 1400kbps. I am obviously still in the training period till the end of today.

 The question is,
(A) If the SNR margin remains stable will my SNR fall maybe not to 6db on such a long line but at least lower than the 12 or 15db showing now

(B) Will my sync speed  increase upwards to around the 1900kbps showing originally or at least rise to somewhere near

 (C) And have i limited that possible speed to only around the 1400kbps showing now , by making a complete bol#*x of the “10 day training period” for my line to find its feet ?.

Thanks in advance.

 New stats are below for your reference.
ADSL Link   Downstream   Upstream
Connection Speed   1408 kbps   448 kbps
Line Attenuation   63.5 db   31.5 db
Noise Margin   10.7 db   15.0 db

For reference there are some download speed tests on 512k,on the max connection before the fault and after the fault
                                                      Date                       Downstream        Upstream             
After fault on max …………    07/11/08 18:29        951.35 Kbps       368.86 Kbps         
Before fault on max………….. 29/10/08 17:14      1424.42 Kbps        354.54 Kbps         
Before fault on 512…………… 28/10/08 20:35       266.09 Kbps          165.15 Kbps         

[roseway]

First you can ignore the 'training period'. All it does is establish a benchmark to compare against in any future fault investigation. It doesn't change the physical connection parameters in any way. What does change the connection parameters is DLM (Dynamic Line Management) and this process continues throughout the life of the connection.

In principle, DLM will lower the previously raised target noise margin over a period of several weeks if the connection is stable and the error rates are sufficiently low. This process isn't well documented, and doesn't often seem to happen (probably because of the error rates). So it could happen in your case, but with such a long line I wouldn't bet on it. BT can manually reset the target noise margin, but they don't like doing it for very good reasons, and I'm sure they wouldn't do it for your line. I think that the best you can realistically expect is to stay stable at the present level.

[dizzy4528]

Thanks for the quick reply 

[roseway]

Sorry it wasn't more encouraging

[dizzy4528]

Would a force of SNR using DMT tool down to say 9db ,providing my line is stable enough bring the speed  up and bras eventually up to the 1.5MB profile.I am using a netgear dg834gt

[roseway]

Yes, it probably would, but there is a bit of a risk that it would lead to instability which would cause DLM to increase the target noise margin further. I suggest that you only make small changes at a time, and then give it a day or two to see how it goes.

[dizzy4528]

Cheers ,will give it a go in small increments.

[dizzy4528]

ADSL Link   Downstream   Upstream
Connection Speed   1408 kbps   448 kbps
Line Attenuation   63.5 db   31.5 db
Noise Margin   10.7 db   015.0 db

ADSL Link    Downstream    Upstream
Connection Speed    1568 kbps    448 kbps
Line Attenuation    63.5 db    31.5 db
Noise Margin    10.3 db    15.0 db

Above are the before(top) and after figures for the 30% reduction on the slider scale on dmt .The rate has risen
but the SNR only dropped .4 ,i was expecting a larger drop .The errors on the dmt page are about the same as pre% drop speeds .
Does this look right or am i missing something.

[roseway]

Unless both sets of figures were taken immediately after a re-sync you can't really reach any conclusion. Even if they were, the noise margin can change quite quickly, so the reported figure isn't necessarily an entirely accurate reflection of the noise margin at the instant of syncing. Your new connection speed will lift you up one level in the IP profile (from 1000 to 1250) if it is sustained, so I would leave it there and see how it goes.

[dizzy4528]

Yet again the voice of reason prevents me from going too far and making  a complete bol*@x .
Cheers again

[roseway]

You're welcome

[kitz]

 

[dizzy4528]

Quick update the speed increase on my ip profile has not kicked in just yet but i aint too worrried as i know from this forum it can take some time.
What i would like to know from you learned gentlmen is , are these errror rates acceptable for my line,or have i pushed the speed too far on my
long line. Just to say the interweb as a whole works well with no problems in page loading or downloading errors. SNR is pretty constant at around 10db 
with very very few spikes and only drops to around 9db at worse in the evening.

Connection Speed     1568 kbps     448 kbps
Line Attenuation    63.5 db    31.5 db
Noise Margin    10.4 db    15.0 db

Previous 1 day time = 24 hours 0 sec
SF  = 5080733
CRC = 298
LOS = 81
LOF = 0
ES  =126
Latest 1 day time = 22 hours 37 min 35 sec
SF  = 4791498
CRC = 326
LOS = 30
LOF = 0
ES  = 244

[roseway]

Those error rates aren't bad at all. 244 errored seconds in 22 hours odd is about 11 ES per hour, which is quite modest.

[dizzy4528]

  Thanx again.