In reply to a post by MrSaffron:

No because, the phone line does not fly like a crow, but wriggles up and down the road

I appreciate that but proportionaly the line will be shorter than someone who is 5km away for the exchange. I guess the question I asked should be - is 24dB a reasonable loss for that distance from the exchange?

5km of cable is around 60dB attenuation

2km is around 24dB of cable

Have a look at http://en.wikipedia.org/wiki/ITU_G.992.1 and http://www.itu.int/rec/T-REC-G.992.1-199907-I/en

9.5.1   Near-end test parameters
The following near-end test parameters shall be provided at the ATU-C and the ATU-R:

o     Attenuation (ATN): The attenuation is the difference in dB between the power received at
        the near-end and that transmitted from the far-end. Received signal power in dBm is the sum
        of all data-carrying (i.e. bi > 0) DMT subcarrier powers averaged over a 1 second period.
        Transmitted signal power is �3.65 �2n + 10 log (sigma gi**2 ) dBm, summed over the data-carrying
        subcarriers. The attenuation ranges from 0 to 63.5 dB with 0.5 dB steps.

o        Signal-to-Noise Ratio (SNR) margin: The signal-to-noise ratio margin represents the
        amount of increased received noise (in dB) relative to the noise power that the system is
        designed to tolerate and still meet the target BER of 10**-7 , accounting for all coding
        (e.g. trellis coding, RS FEC) gains included in the design. The SNR margin ranges from
        �64.0 dB to +63.5 dB with 0.5 dB steps.

If you understand those documents you'll have more understanding than anyone here (speaking for myself of course).

In reply to a post by john2007:

Have a look at http://en.wikipedia.org/wiki/ITU_G.992.1 and http://www.itu.int/rec/T-REC-G.992.1-199907-I/en

9.5.1   Near-end test parameters
The following near-end test parameters shall be provided at the ATU-C and the ATU-R:

o     Attenuation (ATN): The attenuation is the difference in dB between the power received at
        the near-end and that transmitted from the far-end. Received signal power in dBm is the sum
        of all data-carrying (i.e. bi > 0) DMT subcarrier powers averaged over a 1 second period.
        Transmitted signal power is �3.65 �2n + 10 log (sigma gi**2 ) dBm, summed over the data-carrying
        subcarriers. The attenuation ranges from 0 to 63.5 dB with 0.5 dB steps.

o        Signal-to-Noise Ratio (SNR) margin: The signal-to-noise ratio margin represents the
        amount of increased received noise (in dB) relative to the noise power that the system is
        designed to tolerate and still meet the target BER of 10**-7 , accounting for all coding
        (e.g. trellis coding, RS FEC) gains included in the design. The SNR margin ranges from
        �64.0 dB to +63.5 dB with 0.5 dB steps.

If you understand those documents you'll have more understanding than anyone here (speaking for myself of course).

That's what I said, I think
Thanks John, some heavy reading to do now then

Such detail is unnecessary, unless you are a Signals Specialist.

In reply to a post by XRaySpeX:

Such detail is unnecessary, unless you are a Signals Specialist.

I don't see it as detailed. I'm just trying to clear up in my own mind what is happening and unless you have a modicum of understanding then it is very hard to see what is happening. I'm just asking for help in doing so.
For instance I'm still not perfectly clear whether it is better to have a high or low Rx Noise Margin. I'm thinking that a high NM will give you a stable line but at the cost of reduced speed.
The Noise Margin must be a function of the Signal to Noise Ratio which is the only concrete measurable quantity. So I assume that to achive a required margin the speed is reduced to reduce attenuation and therefore increase SNR?
In other word there is a required minimum SNR to support a given data rate. However, because the noise levels will vary a margin is added to the requires SNR to allow for this without compromising stability. This gives you the Recieve Noise Margin which is in fact an adjusted SNR figure to allow for variations in noise.
They are all interelated so an understanding of each perameter and how it interelates with the others is, IMHO, quite important.

A number of misconceptions!

Principally, that attenuation is a function of anything other than the line (length & quality) itself and, possibly, the ADSL mode being employed. Generally, it remains constant from day to day.

Yes, a high NM will give you a stable line but at the cost of reduced speed. But it could be indicative of past instability triggering the DLM imposing a higher Target NM. Or it could indicate you last Synced at night when it was noisy. Or it could be that the ISP is capping you. ... The Target NM is decided by the ISP's DLM at some standard value (like 3, 6, 9 , 12) and applied at Sync time, from which the NM then drifts.

Have a read here: http://www.kitz.co.uk/adsl/linestats.htm It's simpler than those others.

I have been running router stats since Monday and things have been pretty consistent. Mt profile has raised to 5.5Mbps since Wednesday
Router stats shows a fairly consistent..
Synce Speed of 6,848 kbps
RX Noise Margin 14.4dB

With afew exceptions the NM has vaied within +/- 1dB the exceptions are..

Thu 07 Apr 2011 14:23:04 Both Sync Speed and Noise Margin went to 0. The router up time restarted at this point.

Thu 08 Apr 2011 04:19:04 Noise Margin showed a blip down to 7dB.

Thu 07 Apr 2011 05:13:04 Noise Margin showed a blip down to 11dB.

I can't be sure as I wasn't running stats then but the timing of these blips seems to coincide with a similar time last week when my profile was reduced. I may be just imagining this though as I have no stats to prove it.

Is a short complete loss as shown in the first instance above and the couple of minor blips that followed enough to trigger a reduction in profile/increase in target NM?

The profile should only be lowered in response to a poor sync event and it would happen within a few hours. The algorithms for lowering/raising profiles and target SNR margin are not published, so only guess work I'm afraid. Do you have many HEC, CRC, or FEC errors? A lot of those may trigger a rise in target SNR margin.

What is clasified as a poor sync event? Is it a resync that results in a lower sync speed or could a couple of synce over a few hours that resulted in the same sync speed being avhieved be considered as poor?
I am seeing the odd event being graphed where my sync and noise figures go to zero for < 1 minute. My uptime resets to zero at the same time and my router seems to restart. Not sure which is cause and which is effect here . Is the router restating as a result of the poor line conditions or are the line conditions shown as a result of the router restarting??
These have only occured in the last 24 hours (since I normalised my wiring and went back to NTE5 socket with ADSL front plate as my filter. I have been using test socket with a seperate filter up to yesterday. Could it be my faceplate filter? It's brand new.

As for errors. The only stats I can find on my router are collisions, of which there are non.

Thanks and Regards
Clive