[GLLUG] Broadband signal to noise ratio

[Christopher Hunter]

On 18/09/17 08:39, Ken Smith via GLLUG wrote:
>
>
> Chris Bell via GLLUG wrote:
>> Hello,
>> I have a basic FTTC broadband and have noticed that both up and down
>> attenuation figures have remained constant but the signal to noise 
>> ratio drops
>> significantly when the ground is dry, although it soon recovers when 
>> the ground
>> is wet. Is this to be expected? The overall speed reaches the 
>> specified cap at
>> just under 40Mbps, but with many errors logged by the modem, although 
>> I do not
>> see errors in the final downloads.
>> I assume this is due to general radiation such as mobile phone 
>> transmission.
>> When the original lead covered cable was replaced many years ago the 
>> engineer
>> managed to find and re-use the rusty remains of the original iron 
>> pipe laying
>> on the ground and feed the new plastic coated cable through it, but 
>> it was
>> already too short to reach the house. I believe the first phone was 
>> installed
>> when the house was built in 1933, and there was no requirement for 
>> the cable
>> to be laid in a straight line or buried, so it was routed round the 
>> small
>> front garden.
>> The direct route  for a new BT plastic duct from the pavement manhole 
>> at the
>> correct depth would be under crazy paving for much of the way, 
>> although not
>> close to any other services.
>> Thanks for any information.
>
>
> I believe that VDSL follows the same/similar scheme as ADSL where 
> there are many carriers carrying QAM modulated signals. Looking at my 
> VDSL stats (currently running downstream at just under 40 Mbps) its 
> using up to BIN (their language for the carrier number)  2600 which 
> works out to around 11 MHz. (4.3125×2600).
>
> Mobile phone signals are up around 900 Mhz and above so I doubt there 
> is much interference from them.What I have noticed is that dampness 
> does affect the capacity of the line to sync the higher speeds and my 
> VDSL sync rate does vary with the weather. I have seen ADSL services 
> show this behaviour too. Also there is crosstalk between the pairs in 
> the cable bundles in the multicore cables that run from the 
> manhole/phone pole back to the green street cabinets, and I believe 
> this also affects the performance. Is your iron pipe dry inside? The 
> BT cable's capacitance will change if its lying in water especially if 
> there are any nicks in the plastic sheath allowing moisture in.
>
> I'm astonished xDSL broadband works at all over lines originally 
> intended for baseband voice service. Aren't you ex-BBC Chris? I recall 
> hearing that TV OB services sometimes used specially equalised 
> telephone pairs to carry baseband PAL TV (up to 8Mhz) back to the 
> studio where other feeds weren't practical. I was surprised at the 
> time that enough bandwidth was available in a telephone pair to carry 
> PAL TV.
>
> :-) Ken
>
>
I was surprised when I found that some of the TfL CCTV camera signals 
were conveyed for quite considerable distances over twisted pair cable!  
There was a "launch amplifier" at the camera end, which provided some 
amount of pre-emphasis, and a receive amplifier at the instation to 
restore the picture to something normal and feed them into the video 
switching matrix.

Eventually, the twisted pair feeds were supplanted by fibre feeds - all 
TfL's camera feeds now run over fibre and are conveyed as video over 
I/P, routed through cheap commodity routers and switches and with severe 
delays and very degraded quality.  The former analogue system, switched 
by  American Dynamics video matrices provided close to broadcast 
standard pictures, and highly responsive remote control of pan, tilt, 
zoom and focus.  Today's system is a nasty, slow, pixellated mess.

Still - that's progress for you!

Cheers

Chris