I have been noticing for many months that my attenuation (not so much my
SNR) has been varying. As there is a possibility of a speed enhancement (I
am with Plusnet) in the near future I decided to write some code to monitor
the situation. I had been thinking about this for a long time.
My attenuation has been varying from 53 to 58 db which is uncomfortably
close to the old limit of 60 dB. the SNR however is quite good. When I
developed the code (more later) and have been monitoring the situation the
attenuation suddenly dropped to 52.5 and (apart from one sample) has stayed
there ever since! I have now been running the sampling program since 28
February. I did have some earlier results from a few days before that but I
have discarded them as the format was not quite the same.
My sampling program depends on a Freeware package called Teraterm and is
customised for my SAR110 router. It should be relatively easy to customise
for any other router with a telnet interface. The results are displayed in
Internet Explorer (sorry it does not work for Firefox at the moment). The
data can be generated in CVS and/or XML format by simple configuration. The
IE display uses XML data islands.
For those of you interested here are the results to date sampled two to
three times a day.
http://www.soroban.co.uk/routerstats/statistics.htm
A page describing the macro code and how to configure it is
http://www.soroban.co.uk/routerstats. This contains download links for
Teraterm and for the macro code and necessary HTML pages to display the
data.
http://www.soroban.co.uk/routerstats
I had been (possibly still am) interested to see if I could find any
correlation between my varying attenuation and the weather or any other
external factors. I will have to wait until it changes again however to do
that - the jump to 55 db was around the time I needed to reboot my router.
John Steele
jcs AT soroban DOT co DOT uk
"Brian Morrison" <(E-Mail Removed)> wrote in message
news

(E-Mail Removed)...
> Gareth wrote:
>> Does a fluctuating variation in downstream Noise Margin, over a period of
>> a few hours, from 26 db to 14 db suggest a line fault?
>
> Possibly, but probably not.
>
> My SNR margin is typically between 31 and 36 dB. However, sometimes it
> dips somewhat, I've seen values between 7 and 24dB at various times. THe
> modem doesn't lose sync, in fact if I didn't monitor the stats I'd
> probably not know it happens.
>
> The low SNR periods are quite short, the longest I've seen was somewhere
> around 15dB for a few hours but not static during that time.
>
> I think it is probably interference that gets into a phone line pair in
> the same cable and is related to a faulty household item somewhere in the
> two villages served by the exchange.
>
> --
>
> Brian Morrison
>
> please observe reply-to address