Line length / line attenuation correlation query.

I've always thought that as the line length increases then so does the line
attenuation, but not necessarily linearly. However I don't understand why
it is that the upstream line attenuation should be so much less than the
downstream line attenuation since line length must be the same, indeed I
would have thought that the same cables were used for both. I know there
must be a simple explanation, but as I don't seem to be able to understand
this, could someone please explain it in simple terms.

Also, is there a way that I can find out what my actual line length is. I
know in the dim and distant past that a dialling a certain number would lead
to a test being carried out to tell you, but I believe that facility no
longer exists

In my case I believe that the line length is around 2.5 km and my downstream
attenuation is 44.1 dB but upstream attenuation is 23.6 dB. Currently on
ADSL2+ I am synching at 7667 kbps with an SNR that varies from around 9.8 dB
during the day but this drops to low values of around 4.5 dB overnight.
Strangely when connected to ADSL just over a month ago I was synching at
7617 kbps with similar values of line attenuation and SNR so don't appear to
have any benefit in speed for the change to ADSL2+ although my ISP estimated
that I should be getting around 13 Mbps. Does this seem to be as it should
be?
--
Heard melodies are sweet, but those Unheard are sweeter
(E-Mail Removed) FN 2°1° +, Mungo Brandybuck of Buckland

On 10/11/10 01:00, Flyiñg Ñuñ 2°10 + wrote:

> I've always thought that as the line length increases then so does the line
> attenuation, but not necessarily linearly. However I don't understand why
> it is that the upstream line attenuation should be so much less than the
> downstream line attenuation since line length must be the same, indeed I
> would have thought that the same cables were used for both.

Upstream users lower frequencies than downstream. The cable is more
efficient at carrying these lower frequencies. Hence less upstream
attenuation.

Rgds

Denis McMahon

"Flyiñg Ñuñ 2°10 +" <(E-Mail Removed)> wrote:

>I've always thought that as the line length increases then so does the line
>attenuation, but not necessarily linearly. However I don't understand why
>it is that the upstream line attenuation should be so much less than the
>downstream line attenuation since line length must be the same

Attenuation due to line capacitance increases with frequency and the
upstream uses the lowest frequencies.

The reported attenuation is the average of the frequencies in use.

In article <(E-Mail Removed)>, Flyiñg Ñuñ 2°10 +
<(E-Mail Removed)> writes

>In my case I believe that the line length is around 2.5 km

About the same as mine.

> and my downstream
>attenuation is 44.1 dB but upstream attenuation is 23.6 dB.

D:23 U:14

> Currently on
>ADSL2+ I am synching at 7667 kbps with an SNR that varies from around 9.8 dB
>during the day but this drops to low values of around 4.5 dB overnight.

Street lighting? Overhead drop from pole?

SNR here is pretty consistent at 9.9-10.0 dB. If I power cycle the
router it usually starts off at about 12 but settles down to 9.9/10.

>Strangely when connected to ADSL just over a month ago I was synching at
>7617 kbps with similar values of line attenuation and SNR

Think you were doing OK there. Did you try disconnecting any extension
wiring and connecting to the test socket?

> so don't appear to
>have any benefit in speed for the change to ADSL2+ although my ISP estimated
>that I should be getting around 13 Mbps. Does this seem to be as it should
>be?

Run DMT (http://dmt.mhilfe.de) on your router and put a screenshot up
somewhere so we can take a look at the SNR (dB per tone) graph. Do two,
one for when you have 9.8dB SNR and another for when it falls to 4.5.

What router are you using?

--
(\__/)
(='.'=)
(")_(")

Denis McMahon wrote:
> On 10/11/10 01:00, Flyiñg Ñuñ 2°10 + wrote:
>
>> I've always thought that as the line length increases then so does
>> the line attenuation, but not necessarily linearly. However I don't
>> understand why it is that the upstream line attenuation should be so
>> much less than the downstream line attenuation since line length
>> must be the same, indeed I would have thought that the same cables
>> were used for both.
>
> Upstream users lower frequencies than downstream. The cable is more
> efficient at carrying these lower frequencies. Hence less upstream
> attenuation.
>
> Rgds
>
> Denis McMahon

Thank you Denis, what frequency range do they use then? High enough to
incur the skin effect we used to talk about in the old days I presume. I
suppose I should have thought of that.
--
Heard melodies are sweet, but those Unheard are sweeter
(E-Mail Removed) FN 2°1° +, Mungo Brandybuck of Buckland

Mike Tomlinson wrote:
> In article <(E-Mail Removed)>, Flyiñg Ñuñ 2°10 +
> <(E-Mail Removed)> writes
>
>> In my case I believe that the line length is around 2.5 km
>
> About the same as mine.
>
>> and my downstream
>> attenuation is 44.1 dB but upstream attenuation is 23.6 dB.
>
> D:23 U:14
>
>> Currently on
>> ADSL2+ I am synching at 7667 kbps with an SNR that varies from
>> around 9.8 dB during the day but this drops to low values of around
>> 4.5 dB overnight.
>
> Street lighting? Overhead drop from pole?
>
And all the other interference around. Medium wave radio transmissions and
the like. The drop in SNR does star a considerable time before the street
lighting comes on so there are definitely other sources involved.

> SNR here is pretty consistent at 9.9-10.0 dB. If I power cycle the
> router it usually starts off at about 12 but settles down to 9.9/10.
>
When I poweer off my router then the SNR when I reboot it is around 6 dB so
overnight the drop is down to around 0 dB producing dropped connections and
a resetting of the SNR to around 6 dB at that time. This means that when
the daylight hours arrive the SNR ups to 10 dB or sometimes more.
. .
>> Strangely when connected to ADSL just over a month ago I was
>> synching at 7617 kbps with similar values of line attenuation and SNR
>
> Think you were doing OK there. Did you try disconnecting any extension
> wiring and connecting to the test socket?
>
Oh! Yes been through that many times before resulting in Openreach rewiring
my main socket and replacing all the outside cabling to our nearest
telegraph pole which is about 25 feet away. The street cabinet is about 100
feet away from that. I also now have a filtered face plate and the only
other connection is for the 'phone and Sky box. Disconnecting them makes no
difference as far as I can see.
>
>> so don't appear to
>> have any benefit in speed for the change to ADSL2+ although my ISP
>> estimated that I should be getting around 13 Mbps. Does this seem
>> to be as it should be?
>
> Run DMT (http://dmt.mhilfe.de) on your router and put a screenshot up
> somewhere so we can take a look at the SNR (dB per tone) graph. Do
> two, one for when you have 9.8dB SNR and another for when it falls to
> 4.5.
>
Is that the bit loading graph?

> What router are you using?

2wire-2700HGV
--
Heard melodies are sweet, but those Unheard are sweeter
(E-Mail Removed) FN 2°1° +, Mungo Brandybuck of Buckland

Peter Crosland wrote:
> "Flyiñg Ñuñ 2°10 +" <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>> I've always thought that as the line length increases then so does
>> the line attenuation, but not necessarily linearly. However I don't
>> understand why it is that the upstream line attenuation should be so
>> much less than the downstream line attenuation since line length
>> must be the same, indeed I would have thought that the same cables
>> were used for both. I know there must be a simple explanation, but
>> as I don't seem to be able to understand this, could someone please
>> explain it in simple terms.
>>
>> Also, is there a way that I can find out what my actual line length
>> is. I know in the dim and distant past that a dialling a certain
>> number would lead to a test being carried out to tell you, but I
>> believe that facility no longer exists
>>
>> In my case I believe that the line length is around 2.5 km and my
>> downstream attenuation is 44.1 dB but upstream attenuation is 23.6
>> dB. Currently on ADSL2+ I am synching at 7667 kbps with an SNR that
>> varies from around 9.8 dB during the day but this drops to low
>> values of around 4.5 dB overnight. Strangely when connected to ADSL just
>> over a month
>> ago I was synching at 7617 kbps with similar values of line
>> attenuation and SNR so don't appear to have any benefit in speed for
>> the change to ADSL2+ although my ISP estimated that I should be
>> getting around 13 Mbps. Does this seem to be as it should be?
>
> It looks like a straighforward mistake by Plusnet in the estitmation.
> The actual line length is not very relevant other than to show an
> approximate value for attenuation. The faclity did exist on an earler
> variant of 17070 but it is no longer available to the public. In any
> case it was, AFAIK, always a calcutated value based on the
> attenuation. A check here
> http://www.kitz.co.uk/adsl/max_speed_calc.php
>
> shows that the figures you are getting are very close to the
> estimate. Given the line attenuation and noise figures you are not
> going to get a significantly higher figure whatever you do. Stop
> worrying about it!
Not worried about it Peter, just interested in understanding the workings
of it all, which involve answering little queries that I can't work out for
myself. Call it an enquiring mind. The more I ask, the more information I
get and that leads to a better understanding of it all.
--
Heard melodies are sweet, but those Unheard are sweeter
(E-Mail Removed) FN 2°1° +, Mungo Brandybuck of Buckland

Flyiñg Ñuñ 2°10 + wrote:
> I've always thought that as the line length increases then so does the line
> attenuation, but not necessarily linearly. However I don't understand why
> it is that the upstream line attenuation should be so much less than the
> downstream line attenuation

Basically because it uses the lower frequencies.

The whole concept of attenuation is a bit murky..since its not constant
with frequency and broadband..uses..a broad bad of frequencies..

since line length must be the same, indeed I
> would have thought that the same cables were used for both. I know there
> must be a simple explanation, but as I don't seem to be able to understand
> this, could someone please explain it in simple terms.
>
> Also, is there a way that I can find out what my actual line length is. I
> know in the dim and distant past that a dialling a certain number would lead
> to a test being carried out to tell you, but I believe that facility no
> longer exists

My old D-link told me. loop length 10,000 feet it used to say.


>
> In my case I believe that the line length is around 2.5 km and my downstream
> attenuation is 44.1 dB but upstream attenuation is 23.6 dB. Currently on
> ADSL2+ I am synching at 7667 kbps with an SNR that varies from around 9.8 dB
> during the day but this drops to low values of around 4.5 dB overnight.
> Strangely when connected to ADSL just over a month ago I was synching at
> 7617 kbps with similar values of line attenuation and SNR so don't appear to
> have any benefit in speed for the change to ADSL2+ although my ISP estimated
> that I should be getting around 13 Mbps. Does this seem to be as it should
> be?

The crucial thing is recieved SNR. If you are going down towards 4,
that's pretty good, and its unlikely you will get better than that.

The issue with ADSL is not only do the wanted signals get less with
length, but the interference from everything else gets worse.

Which is why speeds plummet over about 2.5Km.

In article <(E-Mail Removed)>, Flyiñg Ñuñ 2°10 +
<(E-Mail Removed)> writes

>And all the other interference around. Medium wave radio transmissions and
>the like. The drop in SNR does star a considerable time before the street
>lighting comes on so there are definitely other sources involved.

Heating system kicking on? Easy enough to check.

>Is that the bit loading graph?

Yes. We'd be looking to see of there are any "notches" (dropouts) in
your frequency graph - that would imply a source of interference on that
specific frequency.

>> What router are you using?
>
>2wire-2700HGV

Supposed to be very good at making the best of marginal lines.

--
Mike Tomlinson

Meanwhile, at the uk.telecom.broadband Job Justification Hearings, Andy
Burns chose the tried and tested strategy of:

> Flyiñg Ñuñ 2°10 + wrote:
>
>> what frequency range do they use then?
>
> Depends if you're ADSL, ADSL2 (nobody in the UK?) or ADSL2+
>
> http://en.wikipedia.org/wiki/File:ADSL2_frequencies.png

I think that image is at least partially bollocks - if ADSL2+ really did use
up to 12MHz, I'd be getting more than ~9Mbps.

--
<http://ale.cx/> (AIM:troffasky) ((E-Mail Removed))
19:28:22 up 9 days, 22:58, 6 users, load average: 0.46, 0.11, 0.06
"I am utterly appalled at how I have been treated like a criminal"
-- Andrew Crossley, ACS:Law, 13 August 2010