Help rqd with odd ADSL fault

My office (in Uxbridge) has been suffering with an intermittent ADSL service
for about three weeks now. A succession of BT engineers have apparently
fixed it; attributing the fault to a variety of causes such as earth on the
line, frozen card at the exchange etc etc, but still the link is up and down
like a yo-yo.

Now for the odd bit - when the link dies, it can usually be persuaded to
reconnect by plugging in an old handset to the base band side of the
splitter and leaving it off the hook for a while after which all will be ok
for a while!

I have a few ideas of my own (bad joint with non-linear conduction
characteristics causing intermodulation distortion - 'off hook' hand set
applies sufficient DC bias to move operating point to a more linear part of
conduction curve?). But, though I am a qualified electronic engineer, I know
very little about telephone systems and anyway surely this type of fault
would show up on the normal line test.

Any ideas?

Nick H

Nick Highfield <(E-Mail Removed)> wrote:
> My office (in Uxbridge) has been suffering with an intermittent ADSL service
> for about three weeks now. A succession of BT engineers have apparently
> fixed it; attributing the fault to a variety of causes such as earth on the
> line, frozen card at the exchange etc etc, but still the link is up and down
> like a yo-yo.
<snip>
> Any ideas?
Replied on uk.telecom.

It's a good idea to crosspost, rather than post multiple times on
multiple groups.

looks like a high resistance fault, where a joint is almost open circuit,
usually due to corrosion.


>
>
>
>

"Ian Stirling" <(E-Mail Removed)> wrote

> It's a good idea to crosspost, rather than post multiple times on
> multiple groups.

Sorry, don't know how.

--
NHH

"TheDragon" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed). ..
> looks like a high resistance fault, where a joint is almost open circuit,
> usually due to corrosion.
>
Would this not show up on normal line test?
--
NHH

Nick Highfield wrote:
> "TheDragon" <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed). ..
>> looks like a high resistance fault, where a joint is almost open
>> circuit, usually due to corrosion.
>>
> Would this not show up on normal line test?

Nope....

Nick Highfield wrote:
> "Ian Stirling" <(E-Mail Removed)> wrote
>
>> It's a good idea to crosspost, rather than post multiple times on
>> multiple groups.
>
> Sorry, don't know how.

Why not find out..?

Ivor

On Mon, 22 Nov 2004 10:33:00 -0000, "Nick Highfield"
<(E-Mail Removed)> wrote:

>
>
>
>
>"Ian Stirling" <(E-Mail Removed)> wrote
>
>> It's a good idea to crosspost, rather than post multiple times on
>> multiple groups.
>
>Sorry, don't know how.

well, dur! Put multiple groups in the to: line...

Well, engineer Dave spent most of yesterday chasing the fault. Apparently we
had far more line loss and noise than would be expected at under a mile from
the exchange, also noise level dropped markedly after 'wetting the line'. He
fiddled with this, tweaked that and just as he thought he had found the
problem, it mysteriously disappeared! Is has to be said that all has been ok
today so fingers crossed.

Thanks to all who offered useful suggestions and for the un-asked for lesson
in Usenet posting ;-)

--
NHH