Testing internal phone wiring - how?

[Andrew Mawson]

Background: BT Infinity with Home Bub 5 installed on 25/02/2016 - speed now

16 plus Mb sec - we are rural and 1.8 miles of copper to the cabinet. This is over double what we were getting - the speed is not the issue.

Connection going up and down like a yo-yo - Hub disconnects then succeeds in re-connecting. Happens frequently - 4 or 5 times at least that day. Report fault to BT and 'Somveer' on the Help Desk has great difficulty understanding the issue, but does a few tests, gets me to isolate my internal wiring at the demarcation point, does more tests and assures me that the fault is with my wiring.

Well I've left my internal wiring disconnected and the problem persists, so obviously it's NOT my wiring. Engineer attending on Monday (01/03/2016), Somveer issuing dire threats that they'll charge me £130 if it's my wiring !!!!

So it occurs to me what tests can I (a fairly well equipped chappy) do to prove my wiring. It's extremely simple in that after the ADSL filter in the BT master socket, all it goes to is my Panasonic DBS Internal Exchange, all phones come off the Panasonic so literally one 'REN' on the line, and it's just a single twisted pair, no ringing circuits to confuse the issue.

The only complication is that the Master Socket is in the house, and the Panasonic Exchange in my office some 50 metres away, and the cable passes through three underground ducts via two buildings on the way and has four Krone Blocks making the joints. (Pucka BT blocks in proper housings) Final Krone block has BT lightning protection arrestors installed.

Is it possible to do a bandwidth test on it? Or maybe a 'loop resistance' test or something along those lines. What would a BT Engineer do to prove someone's internal wiring was out of kilter?

Andrew

[David]

I think all that is moot in the short term.

I think that if at all possible you will have to live without the exchange for a few days and just run the Home Hub off the master socket, and have it set up that way when the engineer calls.

If it is failing in that mode then the main problem is not linked to your wiring.

If the fault clears, then suspect your wiring, but from your post I think you have already shown the fault with the exchange disconnected.

I guess you can/already have put a dumb phone in your office to eliminate the exchange.

Cup of tea and biscuits for the engineer and ask if he can check your (disconnected) wiring? He may well enjoy a look at your setup (assuming he is a proper old style engineer if there are any left).

Oh, the engineer will not be interested in proving that your internal wiring is out of kilter. S/he will test from the master socket with everything else disconnected and if a fault cannot be shown the assumption will be made that it is your wiring at fault.

The best you can do is be set up working directly from the master socket and confirm that the fault shows with that configuration.

Now it is a long time since I had ADSL (Virgin cable here) but I used to use my own ADSL router which had a CLI interface. It recorded loads of statistics on line performance, error rates etc. and these could be used to help diagnose any problems. I don't know if the Home Hub has this capability.

Cheers

Dave R

[Dave Plowman (News)]

Have you looked at the router's log? Any clues there about why it drops out?

[Mike Tomlinson]

En el artículo , Andrew Mawson escribió:

He won't touch it. Keep it simple, just show him that the problem persists when you disconnect your extension wiring at the master socket (demaraction point), with the HH5 connected to the test socket behind the master socket faceplate. You've then shown that it's an issue to do with their side of things.

Hopefully the fault occurs often enough, or you have some way to provoke it into happening, so you can demonstrate it to him.

ps. you /are/ connecting your Panasonic exchange via a filter, yes?

[Andrew Mawson]

These are the last two drop outs - they all look similar to me who doesn't know what they mean:

15:14:30, 27 Feb. (278733.380000) CWMP: session completed successfully 15:14:29, 27 Feb. (278733.150000) CWMP: HTTP authentication success from 15:14:22, 27 Feb. (278725.970000) CWMP: Server URL: Connecting as user: ACS username 15:14:22, 27 Feb. (278725.960000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE' 15:13:59, 27 Feb. (278703.180000) CWMP: session closed due to error: No response 15:13:57, 27 Feb. (278700.980000) CWMP: Server URL: Connecting as user: ACS username 15:13:57, 27 Feb. (278700.970000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE' 15:13:57, 27 Feb. (278700.720000) CWMP: Initializing transaction for event code 4 VALUE CHANGE 13:14:02, 27 Feb. (271506.270000) CWMP: session completed successfully 13:14:02, 27 Feb. (271505.850000) CWMP: HTTP authentication success from 13:13:58, 27 Feb. (271501.990000) CWMP: Server URL: Connecting as user: ACS username 13:13:58, 27 Feb. (271501.980000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE' 13:13:53, 27 Feb. (271497.220000) CWMP: session closed due to error: No response 13:13:53, 27 Feb. (271496.780000) CWMP: Initializing transaction for event code 4 VALUE CHANGE 13:13:41, 27 Feb. (271484.730000) CWMP: Server URL: Connecting as user: ACS username 13:13:41, 27 Feb. (271484.730000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE' 13:13:21, 27 Feb. (271464.780000) CWMP: session closed due to error: No response 13:13:19, 27 Feb. (271462.570000) CWMP: Server URL: Connecting as user: ACS username 13:13:19, 27 Feb. (271462.560000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE' 13:13:18, 27 Feb. (271462.310000) CWMP: Initializing transaction for event code 4 VALUE CHANGE

Andrew

[Andrew Mawson]

I'm assuming that the phone output of the new "SSFP Infinity" socket IS filtered - am I wrong ????

Andrew

[Michael Chare]

There is a line test that you can request possibly by dialing 150. I have done this in the past, but as I no longer have a BT phone I can't try it for you.

[Mike Tomlinson]

En el artículo , Andrew Mawson escribió:

No, that's fine. As you were :)

The log output you quoted isn't very helpful. "Event 4 value change" doesn't tell me much (someone familiar with the HH5 may know more - a google shows many people having the same issue).

Does the router have a 'sync' light that goes out when the line drops?

[Alexis]

You do it the other way, show that it drops out 4-5 times a day with none of your internal wiring connected. So the fault must be with their wiring past the demarc or with the hardware they supplied.

[Andrew Mawson]

Blue light goes out red light comes on until it restarts itself. During this time if you connect to it over the local network it admits to being disconnected, but after a minute or so it sorts itself out and you are back on line.

... 'suppose it's better than the last BT router I had, that had to be 'power cycled' in similar circumstances :)

Andrew

[Andrew Mawson]

Yes, that I'll do, but for my own satisfaction I wanted to profile the internal wiring somehow. I may hook up my Time Domain Reflectometer and see if the various junctions show up in the trace as reflections.

Really need to do a sweep frequency response trace somehow and compare it to a similar length of plain twisted pair 'phone cable - I suppose I could jury rig something with my oscilloscope and a variable oscillator ?

Andrew

[Roger Mills]

I'm not an expert on logs, but those look to me as if your internet session is dropping rather than the line losing sync. In other words, I suspect that ADSL is staying up, and it's just the internet session which is dropping.

Can you log on to the hub and interrogate it about its connection status.

When I had ADSL (I'm now on fibre) my router would tell me what my sync speeds were, and how long the connection had been up.

In another part, it would tell me my WAN IP address, and how long the internet session had been up.

If you can find equivalent items, it will help you to understand what is actually failing.

In addition to using the router's own menus, I had used to run a program called router stats which provided all sorts of useful information - including graphs of what was happening over a period of time. I've just noticed that there's a version for the BT HH5

It's well worth a look. It provided me with some invaluable information with which to beat my ISP over the head when my line kept going up and down like a yo-yo.

[Andrew Mawson]

Roger thanks for that - I've downloaded it and it's capturing data as I type.

The Home Hub does actually say that's it's not connected when the fault arises, then it reconnects after a minute or so.

Andrew

[Alexis]

That isn't necessarily a problem tho, what's on the internal wiring can handle it.

It's not like say TV distribution.

You can certainly hire stuff that does that. How useful it is tho is another matter.

[Roger Mills]

If you can get it to produce a graph of sync speed vs time, it's obvious when the connection drops because the sync speed falls to zero.

I don't know whether the HH version of router stats supports this, but one thing I found incredibly useful was Bitloading. This shows the whole spectrum broken into narrow frequency bands, with the signal to noise ratio of each band - both downstream and upstream. When my line fell over, it was almost always accompanied by a massive loss of SNR in the frequency bands carrying the upstream signal. This forced it to drop the line and re-train, usually re-synching at a snail's pace in the upstream direction. Then it would later recover and decide that it could support a higher speed - causing it to disconnect and re-train again. This could happen umpteen times a day. I printed out loads of graphs in readiness for the OR engineer's visit in case he - or she in my case - couldn't find anything wrong. In the event, she decided that the line between my house and the nearest cabinet was ok but that the line from the cabinet to the exchange was suspect. She swapped this with a spare pair, and the problem went away.

[Brian Gaff]

Surely you need a connection on where it comes in and see if its still all over the place with and without your wiring connected. If it is, then its not your wiring surely? I think this is what you have already done, so just have it ready to show the person when they arrive. Brian

[Andrew Mawson]

Yes that's the plan Brian.

Just out of interest I did put my (ex BT !) Tektronix Time Domain Reflectometer on my internal wiring this morning. Proved total cable length to be 67 metres, and 26 metres to the first Krone block, which oddly gave more reflection than the others further down the line, and still did after I re-terminated the pair, and another 10 metres to the next Krone Block. Ties in with my pacing out !

Only had 2 disconnections so far today :)

Andrew

[Andrew Mawson]

Sorted !

BT man came today, put his tester on the Master Socket, diagnosed a "Rectified Loop" error - probably a damp connection apparently. As cable from pole enters under lounge floor as 'external cable' and the wire to the master socket is 'internal cable' there must be a junction box under the floor' "There's YOUR fault" he says.

Humour me I say - cut the cable at the gutter board and test cable in both directions. Surprise surprise the cable TO THE STREET CABINET shows the fault and into the house is fine !! He then scuttled about at the Street Cabinet (over a mile away) and up at the gutter board, wires the Master Socket again and says 'there you go - fixed' Interrogation revealed he had used a different pair from the Cabinet to the Exchange and a different pair from the gutter board to the Cabinet.

So the ONLY original cable is under my floor and IS OK :) :) :)

Andrew

[Mike Tomlinson]

En el artículo , Andrew Mawson escribió:

Pleased to hear that, thanks for coming back to let us know. Watch you don't get a charge on your bill though - it's not unheard of.

[Capitol]

IME the situation is normal. BT have never spent the money on upgrading the street cabinets and the likely hood is that the next person to play with the cabinet wiring, will disconnect your wires and either forget to put them back properly or change the connections. Many street cabinets are a jungle into which more and more wires have been jammed, so to get at the connections at the back, you have to remove and replace the connections at the front. I feel really sorry for the Open Reach engineers, because the documentation is crap and they get all the flack when yet another connection goes wrong. I get disconnected/ poorly connected at least twice a year and that's in a town with a 100K population. BT also manage to lose my internet connection intermittently during the day, which they deny, but I believe is down to lack of capacity and network switching activity.