Line noise margin variation?

My line noise margin varies between 25 and 14 (or lower).

Does this indicate a faulty line or is variation during the course of a
day - and a few minutes - part of normal line function?

Gareth.

"Gareth" <(E-Mail Removed)> wrote in message
news:41d99943$0$48227$(E-Mail Removed)
> My line noise margin varies between 25 and 14 (or lower).
>
> Does this indicate a faulty line or is variation during the course of
> a day - and a few minutes - part of normal line function?

Its nothing to worry about.

Regards
Sunil

On Mon, 03 Jan 2005 19:13:50 +0000, in article
<41d99943$0$48227$(E-Mail Removed)> "Gareth"
<(E-Mail Removed)> wrote:

> My line noise margin varies between 25 and 14 (or lower).
>
> Does this indicate a faulty line or is variation during the course of a
> day - and a few minutes - part of normal line function?

Typically I have an SNR margin of 33-35dB, but occasionally it falls to
<10dB. The modem trains to line conditions on initial connection and does
not retrain until sync is lost, hence it is possible for changes in line
noise to occur without a retrain, shifting the noise energy into bins that
have multi-bit encodings, and so the SNR margin falls.

If the modem thinks the BER is getting too high it will eventually
retrain when conditions become very poor, but in practice it doesn't
matter much to you provided the line doesn't drop regularly.

--

Brian Morrison

please observe reply-to address

"Brian Morrison" (E-Mail Removed) wrote in message
news(E-Mail Removed)
>
> Typically I have an SNR margin of 33-35dB, but occasionally it falls
> to <10dB.

Jeez.

Highest I've ever seen here is 16dB. Right now, it's 9.5dB. Average is 11dB.
Had another 3 hour outage last night when it stayed below 3dB the whole
time.

*Really* must find a round tuit soon.

On 4 Jan 2005 , in uk.telecom.broadband, "Grant" <(E-Mail Removed)> wrote:

>"Brian Morrison" (E-Mail Removed) wrote in message

>> Typically I have an SNR margin of 33-35dB, but occasionally it falls
>> to <10dB.

>Highest I've ever seen here is 16dB. Right now, it's 9.5dB. Average is 11dB.

Must be a distance thing. Happily I am ~ 30 minutes walk to the exchange
so the figures I have are usually good (and I wish I had some note of the
figures before the BT guy changed the pair back to the exchange, sometime
in December, to be able to compare with the figures today)...


ADSL STATUS (My Plus.Net connection)

Showtime Firmware Version: 2.28b
ADSL Startup Attempts: 1
Elaspsed Time: 0 days 20 hours 7 minutes 35 seconds

SNR Margin 39.8 25.0 dB
Line Attenuation 34.1 21.0 dB
Errored Seconds 0 0
Loss of Signal 0 0
Loss of Frame 0 0
CRC Errors 0 0
Data Rate 576 288 kbps

Only 20 hours as I rebooted it after making some forwarding changes.



ADSL STATUS (My UKFSN connection, using Tiscali's network)

Showtime Firmware Version: 3.27
Startup Attempts: 1
Max Tx Power: -38 dBm/Hz
CO Vendor: ALCATEL_NETWORK
Elaspsed Time: 61 days 15 hours 46 minutes 28 seconds

SNR Margin 29.6 25.0 dB
Line Attenuation 34.1 21.0 dB
Errored Seconds 195 2
Loss of Signal 0 0
Loss of Frame 0 0
CRC Errors 391 2
Data Rate 2272 288 kbps


--------------------------------------------------------------------

For anyone interested, the later firmware for the Conexant / PTI /
DabsValue router (v3.27) has no diagnostic test display within the
browser UI, which is a bit of a shame, IMO. Peter M.



--
PlusNet <http://tinyurl.com/24ymz> - I recommend them and save some cash.

My other ISP : UK Free Software Network <http://www.ukfsn.org>
UKFSN passes all profits to Free Software projects in the UK.

On Tue, 04 Jan 2005 12:15:28 +0000, in article
<(E-Mail Removed)> "Grant" <(E-Mail Removed)>
wrote:

> "Brian Morrison" (E-Mail Removed) wrote in message
> news(E-Mail Removed)
>>
>> Typically I have an SNR margin of 33-35dB, but occasionally it falls
>> to <10dB.
>
> Jeez.
>
> Highest I've ever seen here is 16dB. Right now, it's 9.5dB. Average is 11dB.
> Had another 3 hour outage last night when it stayed below 3dB the whole
> time.

Sorry to hear it, for your information I have a loop attenuation of 42.5dB
currently and the approximate distance from the exchange (and the road
follows the route the cables run in fact) is just under 3.9km as the GPS
measures it. I am very surprised that I get such a high SNR at this
distance, apparently the modems tend to run out of measuring capability at
between 35 and 40dB SNR anyway.

--

Brian Morrison

please observe reply-to address

On Thu, 06 Jan 2005 20:21:36 +0000, Brian Morrison
<(E-Mail Removed)> wrote:

> I am very surprised that I get such a high SNR at this
>distance, apparently the modems tend to run out of measuring capability at
>between 35 and 40dB SNR anyway.

SNR + attenuation is often around 70, if you get 80 then the
interference level must be nice and low. Enjoy !

Phil
--
Splenda - the only sweetener made from chlorine :-)

In article <(E-Mail Removed)>, Phil Thompson
<(E-Mail Removed)> writes
>On Thu, 06 Jan 2005 20:21:36 +0000, Brian Morrison
><(E-Mail Removed)> wrote:
>
>> I am very surprised that I get such a high SNR at this
>>distance, apparently the modems tend to run out of measuring capability at
>>between 35 and 40dB SNR anyway.
>
>SNR + attenuation is often around 70, if you get 80 then the
>interference level must be nice and low. Enjoy !
>
>Phil

So anyone at greater than 7.5km/75db is demonstrably stuffed???

(Hey Phil, I feel we're exchanging similar posts in two different
forums, the other being demon.service, at about the same time...)
Bob

--
robert w hall

In article <(E-Mail Removed)>, Phil Thompson
<(E-Mail Removed)> writes
>On Thu, 06 Jan 2005 20:21:36 +0000, Brian Morrison
><(E-Mail Removed)> wrote:
>
>> I am very surprised that I get such a high SNR at this
>>distance, apparently the modems tend to run out of measuring capability at
>>between 35 and 40dB SNR anyway.
>
>SNR + attenuation is often around 70, if you get 80 then the
>interference level must be nice and low. Enjoy !
>
>Phil


This is a rather limited rule of thumb IMHO Phil.

(I thought it was quite useful until I tried applying it :-))
Looking at a few data for Berkeley and Falfield exchanges, the value for
the sum seems to vary between 75 and 90, and that's a pretty wide range
in SNR.

You _might_ expect there was a correlation of the form
atten + a. SNR = b
where a might be of order 2ish
that might be more serviceable.
--
robert w hall

On Thu, 6 Jan 2005 23:52:48 +0000, robert w hall
<(E-Mail Removed)> wrote:

>Looking at a few data for Berkeley and Falfield exchanges, the value for
>the sum seems to vary between 75 and 90, and that's a pretty wide range
>in SNR.

that is huge, what is the 90 made up of - not 70 attenuation and 20
SNR I hope (envy !)

The "70" came from observing reports from people on longish lines. If
the attenuation exceeds 65 then the inteference level has to be pretty
low to get a decent SNR but this does happen reasonably often.

The time to worry is when you see 50 attenuation and 8 SNR on a 512k
line.

Phil
--
Splenda - the only sweetener made from chlorine :-)