What governs attenuation

Can someone please explain this to me.

If I have a problem with my internal wiring that is affecting my adsl
connection, then when I disconnect the internal wiring completely
should I expect the line attenuation to go up or down?

On Sat, 03 Dec 2005 09:08:35 +0000, GA
<(E-Mail Removed)> wrote:

>If I have a problem with my internal wiring that is affecting my adsl
>connection, then when I disconnect the internal wiring completely
>should I expect the line attenuation to go up or down?

could go either way. Its the noise margin you want to increase.

Phil
--
Tiscali - dialup speeds at Broadband prices, see
http://bbs.adslguide.org.uk/postlist...&Board=tiscali

AOL - the unlimited ISP of choice for heavy downloaders.

On Sat, 03 Dec 2005 09:38:23 +0000, Phil Thompson
<(E-Mail Removed)> wrote:

[snip]
>could go either way. Its the noise margin you want to increase.
>
>Phil


OK - well before disconnecting my internal wiring I had -

42dB attenuation and around 10dB noise margin which recently dropped
to about 4dB.

now I have disconnected the faceplate on my master socket and
connected directly I have -

46.5dB attenuation and around 11dB noise margin.

Looks like I've got a problem with my own wiring but I couldn't
understand why disconnecting it caused the attenuation to go up.

GA

On Saturday, in article
<(E-Mail Removed)>
(E-Mail Removed)se "GA" wrote:

> On Sat, 03 Dec 2005 09:38:23 +0000, Phil Thompson
> <(E-Mail Removed)> wrote:
>
> [snip]
> >could go either way. Its the noise margin you want to increase.
> >
> >Phil
>
>
> OK - well before disconnecting my internal wiring I had -
>
> 42dB attenuation and around 10dB noise margin which recently dropped
> to about 4dB.
>
> now I have disconnected the faceplate on my master socket and
> connected directly I have -
>
> 46.5dB attenuation and around 11dB noise margin.
>
> Looks like I've got a problem with my own wiring but I couldn't
> understand why disconnecting it caused the attenuation to go up.

Possibly impedance matching? But it could be a filtering effect.

If you have a standard dial-up modem lead, you can connect your ADSL box
direct to the internal socket on the Master Socket. No telephone
connected, and an incoming call will mess up the ADSL signal, but it
will eliminate filter problems (or a flaw in the normal ADSL lead).

(This was a method recommended by the Tiscali helpdesk. Despite all the
complaints other people have made about Tiscali, their telephone support
for connection problems is quite good.)

--
David G. Bell -- SF Fan, Filker, and Punslinger.

"I am Number Two," said Penfold. "You are Number Six."

On Sat, 03 Dec 2005 10:36:32 +0000, GA
<(E-Mail Removed)> wrote:

>Looks like I've got a problem with my own wiring but I couldn't
>understand why disconnecting it caused the attenuation to go up.

usually it is a case of the internal wiring picking up interference
and making the higher frequencies unuseable. So the measured
attenuation is at low frequencies which attenuate less.

If you remove the interference the higher frequencies come into play
and they attenuate more.

There might be mucky contacts and some other issues at work too.

Phil
--
Tiscali - dialup speeds at Broadband prices, see
http://bbs.adslguide.org.uk/postlist...&Board=tiscali

AOL - the unlimited ISP of choice for heavy downloaders.

The attenuation along a transmission is a function of the ohmic
resistance of the wires and the impedance, Zo in ohms, of the line.

Mathematically, it is -

Attenuation = R / 2 / Zo decibels per metre of length,

where R is the resistance of the two wires in ohms per per metre,
which itself increases as the squareroot of frequency. Therefore,
four times the freqency doubles resistance and attenuation in
decibels.

Doubling wire diameter halves the wire resistance and attenuation.

The foregoing applies at the higher frequencies, ie., at frequencies
above about 100 kilo-Hertz where Broadband mainly operates.

It is important to distinguish between attenuation and signal loss.

Attenuation is an intrinsic property of the line itself and is fixed.

Signal loss depends on what impedance the line is terminated with.

Ideally, the line should be terminated with the line impedance Zo. If
it is terminated with an impedance higher or lower than Zo then signal
loss will be, in general, greater than line attenuation. But the
difference, in practice, will be relatively small, only a few
decibels.

Thus, the performance of the systen depends mainly on the attenuation
of the transmission line between the local telephone exchange and your
computer.

Not forgetting, of course, the noise picked up by the line. The noise
level also increases with the length of the line. As line length
increases the signal level at the computer decreases and the noise
level increases. With increasing line length the two levels converge
towards each other. When they are both around the same level the
system stops working altogether.

Noise level not only increases with line length, it increases with the
number of working Internet lines in the same cable or otherwise
running in parallel. Interference between circuits running parallel to
each other is known as Crosstalk.

So as traffic expands and the number of customers increases the
quality of service provided inevitably gets worse. No doubt many of
you have noticed this deterioration with time. But you can't expect
service providers to turn away new customers just to maintain a high
grade of service to existing ones. Or can you?
----
Reg.

On Sat, 3 Dec 2005 22:03:18 UTC, "Reg Edwards"
<(E-Mail Removed)> wrote:

> Doubling wire diameter halves the wire resistance and attenuation.

That's new. I always thought doubling the cross sectional area halved
the resistance.

(yes, I do know about skin effect, but that ain't it)
--
[ 7'ism - a condition by which the sufferer experiences an inability
to give concise answers, express reasoned argument or opinion.
Usually accompanied by silly noises and gestures - incurable, early
euthanasia recommended. ]

"Bob Eager" <(E-Mail Removed)> wrote in message
news:176uZD2KcidF-pn2-(E-Mail Removed)...
> On Sat, 3 Dec 2005 22:03:18 UTC, "Reg Edwards"
> <(E-Mail Removed)> wrote:
>
> > Doubling wire diameter halves the wire resistance and attenuation.
>
> That's new. I always thought doubling the cross sectional area
halved
> the resistance.
>
> (yes, I do know about skin effect, but that ain't it)
> --
===========================================

Have another think. It is skin effect which CAUSES attenuation to
halve when wire diameter is doubled.

Twice the wire diameter means twice the surface area of the wire and
half of the attenuation in decibels.

The cross-sectional area of the wire does not enter the argument.
Regarding current flow at high frequencies, the wire is just a
thin-walled tube.

Now concentrate on the jist of my message - when the noise and signal
levels at the receiving end of the line approach each other with
increase in line length, communications break down.

Which is rather an obvious conclusion. But we must not allow somebody
who has heard of skin-effect from cheap magazines to interfere with
genuine elementary education.
----
Reg.

On Sun, 4 Dec 2005 06:01:50 +0000 (UTC), "Reg Edwards"
<(E-Mail Removed)> wrote:

>Have another think. It is skin effect which CAUSES attenuation to
>halve when wire diameter is doubled.

however you said "Doubling wire diameter halves the wire resistance
and attenuation".

is that correct for resistance.

Phil
--
Tiscali - dialup speeds at Broadband prices, see
http://bbs.adslguide.org.uk/postlist...&Board=tiscali

AOL - the unlimited ISP of choice for heavy downloaders.

On Sun, 4 Dec 2005 06:01:50 UTC, "Reg Edwards"
<(E-Mail Removed)> wrote:

> "Bob Eager" <(E-Mail Removed)> wrote in message
> news:176uZD2KcidF-pn2-(E-Mail Removed)...
> > On Sat, 3 Dec 2005 22:03:18 UTC, "Reg Edwards"
> > <(E-Mail Removed)> wrote:
> >
> > > Doubling wire diameter halves the wire resistance and attenuation.
> >
> > That's new. I always thought doubling the cross sectional area
> halved
> > the resistance.
> >
> > (yes, I do know about skin effect, but that ain't it)
> > --
> ===========================================
>
> Have another think. It is skin effect which CAUSES attenuation to
> halve when wire diameter is doubled.

At high frequencies. But you never mentioned that. And it also depends
on the frequency. Yours was a blanket statement.

> The cross-sectional area of the wire does not enter the argument.
> Regarding current flow at high frequencies, the wire is just a
> thin-walled tube.

Of course. But you never said that. And the thickness of the 'tube'
depends on the frequency.

> Which is rather an obvious conclusion. But we must not allow somebody
> who has heard of skin-effect from cheap magazines to interfere with
> genuine elementary education.

No need to descend to cheap shots when you have no idea whether or not I
am qualified in the field (I am). Actually, no need to descend to
personal attacks anyway.

--
[ 7'ism - a condition by which the sufferer experiences an inability
to give concise answers, express reasoned argument or opinion.
Usually accompanied by silly noises and gestures - incurable, early
euthanasia recommended. ]