BT Master Socket - can NTE5 extensions interfere with ADSL?

[davelx]

No, you were not imagining it, there is (or was) a site that gave the distance - I distinctly remember it telling me I was 770m from the exchange.

I'll see if I can find it internally.

 

[plugwash]

But do not use the third pair in the cable for ADSL when the other two pairs are carrying phone signals.

It has spent the entire distance from the exchange running alongside phone signals, doing it for a little bit longer isn't going to harm it.

Do use good quality wiring (I'd reccomend either BT spec phone cable or cat5) for any ADSL stuff though. At a friends house recently him and I replaced some cheap flat cable with cat5 and got an extra megabit per second of speed!

I would try the following in order
1: disconnect the ringer wire from any unfiltered signals.
2: eliminate any branches in the unfiltered signal by moving filters to the branch points
3: reduce the system to a single filter
4: replace any cheap cr*ppy wire in the unfiltered signal with either BT spec cable or cat5

 

[Paper Spark]

No, you were not imagining it, there is (or was) a site that gave the distance - I distinctly remember it telling me I was 770m from the exchange.

I'll see if I can find it internally.

There still is clicky>>

Loads of other info as well.

 

[c128]

There still is clicky>>

Loads of other info as well.

But not distance from the exchange...or, at least, there isn't for the phone number I'm putting in. There's lots of info about the exchange, mind...

 

[Paper Spark]

There still is clicky>>

Loads of other info as well.

But not distance from the exchange...or, at least, there isn't for the phone number I'm putting in. There's lots of info about the exchange, mind...

Use your postcode, then click on the BT ADSL button. It should tell you the straight line distance at the bottom of the page.

 

[bernardgreen]

But do not use the third pair in the cable for ADSL when the other two pairs are carrying phone signals.

It has spent the entire distance from the exchange running alongside phone signals, doing it for a little bit longer isn't going to harm it.

In the street cabling all wires are balanced pairs resulting in minimum cross talk between diferent phone lines.

In the house wiring there is the loop pair ( pins 2 and 5 ) which are balanced and the ringing wire (pin 3 ) which has no wire to balance it meaning the third pair is adjacent to an un-balanced signal wire ( the bell wire ). This is why micro filters have a bell capacitor instead of the much cheaper solution of adding pin 3 for the bell wire into the plug. The capacitor in the micro filter means the balance in the cable is better as the bell wire does not have to be used which whic create the un-balance.

 

[c128]

Use your postcode, then click on the BT ADSL button. It should tell you the straight line distance at the bottom of the page.

Aha! Thanks - it shows the distance for a post-code, but not a line number. The distance is 3.48km - about 2.2miles then.

 

[plugwash]

In the house wiring there is the loop pair ( pins 2 and 5 ) which are balanced and the ringing wire (pin 3 ) which has no wire to balance it meaning the third pair is adjacent to an un-balanced signal wire ( the bell wire ).

I still can't see this making much difference, the ringer signal is way outside the DSL frequency bands and besides its already in the DSL signal.

This is why micro filters have a bell capacitor instead of the much cheaper solution of adding pin 3 for the bell wire into the plug. The capacitor in the micro filter means the balance in the cable is better as the bell wire does not have to be used which whic create the un-balance.

I thought the reason micro filters used a ring capacitor was that is was easier to do that than to filter a third wire seperately. Afaict when ring wire is a factor its because its acting as an antenna attatched to the line.

EDIT: added a missing quote tag

 

[Roo]

This will give you an appox distance bearing in mind your line probably doesn't run as the crow fly's

http://www.dslzoneuk.net/checker.php

 

[TicklyT]

In the house wiring there is the loop pair ( pins 2 and 5 ) which are balanced and the ringing wire (pin 3 ) which has no wire to balance it meaning the third pair is adjacent to an un-balanced signal wire ( the bell wire ).

I still can't see this making much difference, the ringer signal is way outside the DSL frequency bands and besides its already in the DSL signal......

For POTS (Plain Old Telephone Systems), the capacitor in the ringer circuit in an NTE 5 was designed just to block DC. in effect a 'high pass filter' passing signals at frequencies above about 10 hertz. to the phone ringers via pin 3 of the connector. Now we have broadband, that capacitor will pass the DSL signal (and any unwanted signal reflections or pickup on the unbalanced ringer wire back onto the balanced pair carrying the DSL signal)

In a DSL filter, the signal is filtered before it reaches the ringing capacitor, so it behaves as a band pass filter, blocking both the DC component and the DSL signal from it's locally generated ringer connection.

So disconnecting the now redundant ringer wire will reduce the signal attenuation and noise on the DSL.

 

[plugwash]

In the house wiring there is the loop pair ( pins 2 and 5 ) which are balanced and the ringing wire (pin 3 ) which has no wire to balance it meaning the third pair is adjacent to an un-balanced signal wire ( the bell wire ).

I still can't see this making much difference, the ringer signal is way outside the DSL frequency bands and besides its already in the DSL signal......

For POTS (Plain Old Telephone Systems), the capacitor in the ringer circuit in an NTE 5 was designed just to block DC. in effect a 'high pass filter' passing signals at frequencies above about 10 hertz. to the phone ringers via pin 3 of the connector. Now we have broadband, that capacitor will pass the DSL signal (and any unwanted signal reflections or pickup on the unbalanced ringer wire back onto the balanced pair carrying the DSL signal)

In a DSL filter, the signal is filtered before it reaches the ringing capacitor, so it behaves as a band pass filter, blocking both the DC component and the DSL signal from it's locally generated ringer connection.

So disconnecting the now redundant ringer wire will reduce the signal attenuation and noise on the DSL.

right but i can't see a ringer wire for a different line (or the same line after filtering) running on a different pair in the same cable causing any significant interference.

 

[JOHNBOY42]

If you say the signal is ok at the nte5,then work from there,,,,,,,,,,take off the bell wire as already suggested .check to two extension sockets and make sure none of them are not master sockets with a capacitor in it,the old type of master sockets are the same size as the extension sockets and you will have no problem with speech but if is has a capacitor in it this will affect your dsl signal.......are the cables running across or parallel to a mains cable as the dsl signal will pick up electrical interference ......what type of wiring is the extensions run off.....if the wiring appears to be fault free then it could be the quality of it..as them cheap nasty extensions cause the dsl signal to fall at the first metre.........if a re-wire is required dont use anything less than cat5 and star wiring is out of the question..just run in series.............
I know you have tested in the Bt master socket and got a good signal but for how long??

 

[TicklyT]

In the house wiring there is the loop pair ( pins 2 and 5 ) which are balanced and the ringing wire (pin 3 ) which has no wire to balance it meaning the third pair is adjacent to an un-balanced signal wire ( the bell wire ).

I still can't see this making much difference, the ringer signal is way outside the DSL frequency bands and besides its already in the DSL signal......

For POTS (Plain Old Telephone Systems), the capacitor in the ringer circuit in an NTE 5 was designed just to block DC. in effect a 'high pass filter' passing signals at frequencies above about 10 hertz. to the phone ringers via pin 3 of the connector. Now we have broadband, that capacitor will pass the DSL signal (and any unwanted signal reflections or pickup on the unbalanced ringer wire back onto the balanced pair carrying the DSL signal)

In a DSL filter, the signal is filtered before it reaches the ringing capacitor, so it behaves as a band pass filter, blocking both the DC component and the DSL signal from it's locally generated ringer connection.

So disconnecting the now redundant ringer wire will reduce the signal attenuation and noise on the DSL.

right but i can't see a ringer wire for a different line (or the same line after filtering) running on a different pair in the same cable causing any significant interference.

The point is the now redundant ringer wire is split from the speech pair before the filter by the NTE5, and may run all the way round the house. The noise cancelling characteristic of twisted pairs is negated as there isn't a balanced signal on both wires in the pair. The telephone cabling may not even be twisted pair construction anyway. A 4-wire twisted quad construction used to be common practise.

Imagine fitting a 2-way adapter to your television aerial lead. Connect a random length of unscreened wire, or a single wire in a cat 5 cable to the spare coax core, via a 2 microfarad capacitor if you like, and run that wire all the way around your house passing through every room...... How's your TV reception now?

 

[Softus]

TicklyT and bernardgreen (particularly), given that many modems can be encouraged to provide (on a software interface) some figures about line attenuation and noise, can any of this information be used to measure the relative effect of the changes you've been advocating?

 

[bernardgreen]

TicklyT and bernardgreen (particularly), given that many modems can be encouraged to provide (on a software interface) some figures about line attenuation and noise, can any of this information be used to measure the relative effect of the changes you've been advocating?

The modem can only provide an indication of the signal level that it is receiving IF it has that capability. It cannot provide information about attenuation without knowing the signal level sent from the exchange.

The copper loop to the house was designed for nothing more than DC and speech frequencies up to 4 Khz. Getting 2000 times that frequency down the line is a bonus. To do so is part science, part luck and a lot of balck art.

The rule of thumb is to give the ADSL signal the best route you can.

Where a change of cable occurs there is a change of charactoristic impedance CI , CI is a complex impedance of capacity, resistance and inductance and at each change of CI some signal is reflected backwards. The bigger the change the more is reflected.

So minimising the changes in CI reduces loss by reflection.

A stub cable ( that branches from a main route ) that has a dead end can totally reflect the signal back the point where it divided from this gets back to the junction point slightly later than it left and so is interference to the signal that followed it.

Separate the ADSL from the speech / analogue modem signals as close to the point of entry as possible. Some NTE 5 filters have 2 filters. One to keep ADSL awy from phones and second one to reduce the effect of changing CI and reduce speech frequency signals on the ADSL to the computer / hub.

How effective ? A friend was getting less than one Megabyte/second. By taking a single pair from his master socket to a second NTE5 by the computer and wiring all phones to that NTE 5 phone side the rate went up to around 3 Meg / second.