Phone line extension help

[rizred]

Hello everbody,

I seem to have caused a "discussion"!!

Just to confirm the requirement and attempted solution:-

The new "office/ shed/ room" is intended to be used occasionally, and only needs an extension of the main phone line. We have put in a cat5e cable to run from an ethernet port on the existing router to provide internet out there.

The line we have put in, and connected directly on to the BT socket, works fine - ie we get a usable phone line. The problem comes when, having plugged a new phone into the new socket, the internet through the original router does not work.


Would i be right in thinking from your replies that i need a "filtered" socket on the end of the new extention as this is causing some kind of reflected signal?

Is this covered by "master/ secondary" phone sockets? (I cant remember what we used!)


Thanks again!!

D

 

[TonyW2]

All telephone sockets ,with an instrument connected need a filter on them regardless of where they are situated Therefore the shed needs one and the master socket that the router is plugged in to should have one.
Suggest you try this simple solution first.
( The alternative suggestions with filtered master points are far neater but have the same affect.)

 

[Simples]

OK, this is my area of expertise! I am a telephone engineer of 30 years experience. I agree with everything so-far submitted, however I urge you to remove the face plate of the extension socket and physically remove, ie cut-out any yellow capacitor you find on the back of the socket.

The capacitor is only required on the first socket in the premises, aka BT master socket. It's role is to generate line current to power the ringers in the telephones around the house. Having a capacitor on any other extension stop any phone plugged into the extension ringing. This will also kill off any broadband signal via that extension.

Hope you have better luck with this than I am having with my alarm system!!!

 

[rizred]

Hi Simples,

thanks, i'm pretty sure that the socket i installed did have a yellow capacitor - i'll pass on your suggestion and see what happens

Thanks

 

[TonyW2]

Sorry--ignore please.

 

[bernardgreen]

Having a capacitor on any other extension stop any phone plugged into the extension ringing.

Only if the wiring is incorrect on terminals 2 and 5 ( crossed ) as the two caps would be in parallel giving more power to the ringers.

This will also kill off any broadband signal via that extension.

Not just on that extension

Separating ADSL wiring from internal telephone wiring by filtering at the NTE5 is by far the best way to improve Broadband speeds.

 

[drgl]

OK, this is my area of expertise! I am a telephone engineer of 30 years experience. I agree with everything so-far submitted, however I urge you to remove the face plate of the extension socket and physically remove, ie cut-out any yellow capacitor you find on the back of the socket.

The capacitor is only required on the first socket in the premises, aka BT master socket. It's role is to generate line current to power the ringers in the telephones around the house. Having a capacitor on any other extension stop any phone plugged into the extension ringing. This will also kill off any broadband signal via that extension.

Hope you have better luck with this than I am having with my alarm system!!!

Your a phone engineer of 30 years? More than one capacitor on the line will cause ZERO problems, take a look in any SIN complient filter and you will notice they have a ring capacitor to generate the ring voltage "locally". More than one master socket will NOT "kill off" any broadband (ADSL) signal either. If 2 & 5 are crossed on one extension then the phone will ring continuously when plugged in.

Back to the problem :-

"The new "office/ shed/ room" is intended to be used occasionally, and only needs an extension of the main phone line. We have put in a cat5e cable to run from an ethernet port on the existing router to provide internet out there.

The line we have put in, and connected directly on to the BT socket, works fine - ie we get a usable phone line. The problem comes when, having plugged a new phone into the new socket, the internet through the original router does not work. "


So you have a CAT5e cable that runs to the shed, this is plugged into the routers ethernet port so no speed issues or signal reflections( ). You then have a seperate phone line extension that is run off the NTE5. If the router drops the connection when a phone is plugged in there then all you need is :-

(1) A filter :-http://www.bttorj45.com/BTADSLfilter.html

OR

(2) A filtered socket in the shed:- http://www.bttorj45.com/BTsocketmasterADSLfiltered.html

OR

(3) A filtered faceplate on the NTE5 :-http://www.bttorj45.com/BTNTE5ADSLfaceplate6way.html

Option 3 is best if you have the modem plugged into the NTE5 at present? This option does away with any plug in filters. Option 1 will mean everything you currently have will just work!

 

[Simples]

Hi Drgl,

I am speaking from (painful) experience, having made exactly this error on a phone system installation
To explain why this is an issue, you have to remember that two capacitors in parallel work the same as two resistors in series. Ie; the overall capacitance is the Sum of the two capacitors.

Thus two master sockets = double the line capacitance. The line impedance is therefore significantly increased.

Now consider basic line transmission theory. Where there is an impedance mismatch between two networks there will be a 3dB loss of power. 3dB = half power.

The broadband signal broadband signal has already been depleted by line characteristics from the local exchange, it sure don't need this extra hurdle!

By the way, I fully agree with your comments relating to the CAT5e leads. I use Cat6e between my PC and the modem.
Keep up the good work!

 

[drgl]

That doesn't explain how BT state upto 4 filters(with capacitors inside so in essence the same as a master(PABX) socket) are fine?? I've got two masters at home and have no problems. When you say you had this problem on a phone system installation was it a PBX? Many digital PBX's do not "like" PABX sockets and function only with secondary sockets.

 

[bernardgreen]

The filter at the shed end of the cable may improve the problem by absorbing the ADSL signal that has gone down that cable and prevemting it from being reflected. Any signal that is reflected back will arrive at the modem after a delay and therefore out of phase with the incoming signal thus creating distortion of the signal. But by absorbing some of the signal it has reduced the signal level at the modem thus reducing the efficiency of the Broadband linkage.

That is why it is better to avoid having any spur ( or side turning ) in the wiring that the ADSL signal can get into.

Putting the filter at the house end of the cable to the shed will prevent ADSL signal getting onto that length of cable so [1] more signal goes to the modem and [2] reflections with significant delays will not occur and the modem will not have to cope with signals significantly distorted by delayed reflections.

Not the case here but on a weak ADSL signal distortion caused by reflections delayed by a few feet of spur can be the difference between working ADSL and no working ADSL...

The use of a intentional added spur cut to the length that returns a reflection exactly 180 degrees out of phase with the incoming signal is a common method of reducing unwanted signals along co-axial cables to almost remove them. It is refered to as a tuned stub. A similar effect is accidently possible with ADSL on a twisted pair except the system will try to compensate by altering it's parameters to minimise the effect of the accidently created stub.

 

[bernardgreen]

That doesn't explain how BT state upto 4 filters(with capacitors inside so in essence the same as a master(PABX) socket) are fine??

The micro filters only take the line pair ( terms 2 and 5 ) and not the bell wire ( term 3 ) The capacitor is connected to term three of the filters socket so only a phone with a bell connection in its plug will bring the capacitor into circuit.

The ringing voltage ( AC 70 volts ) source at the exchange can feed 4 standard ringers before the AC current becomes high enough to risk a false Off Hook indication. Whether the current is supplied via 1 common capacitor ( in the master socket feed bell wire term 3 ) or four individual capacitors in filters or the telephones themselves is not critical When a telephone plug has only two pins in use that telephone has its own internal capacitor for the ringer. A standard ringer has a REN of 1

Many digital PBX's do not "like" PABX sockets and function only with secondary sockets.

The Off Hook detection in some PABX's is designed to work reliable only with a REN of 2 ( some only one ).

Modern PABX systems tend to use separate pairs for every telephone and they create "extensions" to phones only in software with the PABX configuration and not by having more than one phone per line from the PABX

 

[Simples]

That doesn't explain how BT state upto 4 filters(with capacitors inside so in essence the same as a master(PABX) socket) are fine?? I've got two masters at home and have no problems. When you say you had this problem on a phone system installation was it a PBX? Many digital PBX's do not "like" PABX sockets and function only with secondary sockets.

A good question ...

The capacitors built into the microfilters are typically 22nF (0.000000022pF) and are part of a band pass circuit. As such they are designed, along with coils to present an impedance of 600 Ohms back to the telephone line whilst passing the voice component of the line (300 - 4000 Hz.).

Compare this to the capacitor built into the master socket. This is rated @ 1.8mF (0000018F). It is by comparison far larger and has no line balancing circuitry associated with it.
Thus sticking two in Parallel is going to cause some major issues for a broadband signal. The microfilter is designed to operate against an input capacitance of 1.8mF, instead you present it with 3.6mF

The impedance matching circuitry in the microfilter is now severely compromised and signal loss will result.

You rightly pointed out that up to four microfilters can be connected across one line. That is 4 X 0.22nF = 0.88nF Still less than 1.8mF
Add one more in and the result is 5 X .22 = 1.1mF, and the signal is starting to struggle.

Hope this explains the issue.

 

[bernardgreen]

The capacitors built into the microfilters are typically 22nF (0.000000022pF) and are part of a band pass circuit.

Correct.

As such they are designed, along with coils to present an impedance of 600 Ohms back to the telephone line whilst passing the voice component of the line (300 - 4000 Hz.).

Not quite correct. The circuit is to present a high impedance to the high frequency ADSL signal to prevent it entering the phone.

Compare this to the capacitor built into the master socket. This is rated @ 1.8mF (0000018F). It is by comparison far larger and has no line balancing circuitry associated with it.

Its main purpose is to couple the AC ring voltage from wire 2 to wire 3 which goes to the ringer in the phone(s). The return of the bell is on wire 5.

The value of 1.8uF ( microFarad ) will supply enough AC power to adequately ring 4 standard ringers (total REN = 4 )

This capacitor in a master socket has a second function in line testing. The exchange routiner ( testing equipment ) measures the time taken to charge the capacitor via the 470 ohm resistor in series with the line resistance to obtain an idea of line loop resistance and quality.

In microfilter the capacitor is smaller than 1.8uF as it only has to supply one standard ringer.

You rightly pointed out that up to four microfilters can be connected across one line. That is 4 X 0.22nF = 0.88nF

The effective parallel capacitance presented to the line is far less than the parallel of the four capacitors as they are not in directly in parallel as one end of each capacitor is connected to wire 2 but the other end is connected to pin 3 of the phone and then on to wire 5 via the high impedance of the ringer circuit in the phone if it is a three wire phone.

If it is a two wire phone there is a ring capacitor in phone but this capacitor connects to wire 2 AFTER the band pass filter so does not affect the ADSL signal which the filter has prevented from reaching the phone

 

[Infidel]

Well this thread has been fun to read I think the OP is thoroughly confused by now.
But for all you newbies to telecommunications here's how it used to look.
http://www.samhallas.co.uk/repository/n_diagrams/popular.htm
N 846 I suggest, not a lot has changed as far as basic circuitry is concerned apart from component size, or N 4507 for the complicated stuff

 

[Paul_C]

N 846 I suggest, not a lot has changed as far as basic circuitry is concerned apart from component size, or N 4507 for the complicated stuff

As far as the ringer circuitry is concerned, it is rather different. The old GPO method as depicted in the N-diagrams of the 746 and the various extension plans was to use a single capacitor in the main instrument with 1000-ohm ringers wired in series. If an extension phone had a bell cut-off switch, it was wired to short out the ringer, and where jacks were used for connecting extension phones, shorting contacts were provided to maintain the series bell circuit when the jack was not in use.

The method adopted when the BT431 jacks & plugs were introduced in the early 1980's is to use a single capacitor at the master jack which feeds parallel-connected ringers of much higher impedance - Around 4000 ohms.