Can Copper problems upstream of cab cause VDSL problems?

We have had problems with our overhead since we moved here 7 years ago - it finally broke just after we had BT Infinity installed last summer and the overhead length to the roof line was replaced, and jointed to an old length down the wall which was left untouched (difficult to replace you see as it disappears behind a porch).

All was great, until a few months ago when we started to have problems again with intermittent noisy phone calls, but also VDSL drop outs. Then just over a week ago the noise got chronic and our profile got borked down from 65Mbps to 42Mbps.

So reported the fault online and the line checker actually detected something and offered an appointment which is in for next Wednesday. All set I thought - we'll get the whole cable replaced this time as it's bound to be the wall cable....

However this week BT appear to have replaced the trunk copper cable up the hill from the cab towards the exchange (vans with cable drums, jumper wire on pavement, disturbed pavement lids etc.), and despite several Quiet Line tests I can no longer detect any noise problem on voise calls.

My profile is still stuck at 52Mbps though

Now, the question is, could a problem with the trunk cable on the exchange side of the Fibre Cab actually have caused my VDSL disconnects and profile disruption? I had thought that this must have pointed towards my house connection rather than a problem with the copper on the other side of the cab....

I also need to know whether I should maintain the appointment on Wednesday, or am I risking a NFF charge if in fact they have already fixed the problem by virtue of the new trunk cable over the past few days?

Finally, can I persuade them to reset my profile if the trunk cable was the reason for it getting borked in the first place?

Many thanks for any advice!

Edited by deleted (Sat 18-May-13 18:44:05)

Now, the question is, could a problem with the trunk cable on the exchange side of the Fibre Cab actually have caused my VDSL disconnects and profile disruption?

No.

Post deleted by RobertoS

In reply to a post by Zarjaz:

No.

I know I'm asking to be shot down doing this, but here goes...

I know there's a low-pass filter of some kind on the exchange-side of the tie cables within the FTTC cab, and this is what ensures that only the voice signal passes from the upstream cable into the FTTC cab electronics.

Therefore I understand that there's no way a proper signal can get through the block, and interfere with the VDSL2 signal.

But...

Is there *no* type of fault on the upstream line that could interfere with the downstream line?

I would have thought that some kind of intermittent contact between the A and B lines could have an effect, perhaps caused by water damage.

Surely the loop start / ground start signalling gets passed through the cabinet, otherwise how does the switch know when you pick the phone up?

The damage to the trunk cable was almost certainly due to damage / water ingress from gas main works that took place last year - I remember seeing the exposed BT cable with some nasty bashes in it at the time (probably from jack hammers). If the insulation between legs on a pair has been damaged then surely that it going to affect the performance of the whole line even on the other side of the cabinet, and thus its capability to transmit DSL signals?

In reply to a post by Colin_London:

Surely the loop start / ground start signalling gets passed through the cabinet, otherwise how does the switch know when you pick the phone up?

When you go off-hook, the resistance drops from something like 10megohms to 1000 ohms or less. The exchange detects this by the current flowing in the line.

The filter doesn't prevent this from happening; it only adds components that makes sure any frequency above normal speech gets blocked out.

In reply to a post by Colin_London:

even on the other side of the cabinet, and thus its capability to transmit DSL signals?

That's the question.

I'm good at knowing what the software should be doing, and have a grasp on what goes on in the copper when it is working. But I'm far from skilled in knowing what happens in the copper in fault situations, or what all the fault situations even are. Zarjaz is the man for that part...

I would have thought that water ingress could indeed cause voltage problems, resistance problems, and shorting problems... and that all of those could get transmitted onto the D-side. But I could be wrong.

If the VDSL Cab had something like a PABX line card in it, and effectively 'bridged' two separate phone lines on either side of the cabinet then I would agree that there could be no interaction between a fault on the E side of the cab and the D side.

But I don't think they are isolated from each other by the cab in that way, and you still have a transmission line at VF and below whose DC performance can be affected by a fault at any point?

Zarjaz is the man for that part...

I admire your confidence. !

A damaged cable will give you audible noise, from the resulting earth and battery contacts, low insulation, HR's etc. If this damage is on the E-side, then it cannot be affecting the transmission of the VDSL signal down the D-side.

I wonder if the OP has unlocked his modem to look at attainable rates ? The audible noise could well be from a mullered E-side, but his drop in sync rate is going to be something D-side related, cross talk or some such.

Given the general size of components used in such filters (and the space available in cabinets) , I suspect that they are all High (ADSL/VDSL) - Pass/Reject designs, rather than Low (Audio)- Pass/Reject designs.

For Audio Filters, it is close to Henries and micro-Farads that are required, implying very significant physical sizes.

For VHF/UHF Filters, the values are in the micro-Henry and pica-Farad range, resulting in very small physical sizes, as presently demonstrated in the domestic ADSL/Phone Splitters.

Hence in the Splitters, it is VHF/UHF Reject filters in the Phone leg, whilst the ADSL leg is "straight-through", ie no filtering.

It may be possible to demonstrate the likelihood of problems of the nature described, by shorting the appropriate pins on the Phone side of such Splitters.

Picure of the filters in a cabinet - post #72
ECI Cabinet filters

I've seen the other responses to the thread.

I too believe that it is possible especially if there is a low impedance fault - not quite a short circuit but conduction through water or deposits on terminals. That could drastically alter the load impedance of a filter and totally change the characteristics. However, without seeing the design of the filters in the cabinets it will be hard to say definitely.

An unequivocal NO is certainly not the right answer

Surely the loop start / ground start signalling gets passed through the cabinet, otherwise how does the switch know when you pick the phone up?

that's DC, even the lowest of low pass filters should let that through.

An unequivocal NO is certainly not the right answer

But it is 99.9% right. Nothing is ever completely cut and dried when it comes to ADSL/VDSL faulting. In 2.5 years of working on this stuff, even when a line fails it's PQ test due to an E-side, and it is duly swapped out, this has never made an iota of difference to the results on the VDSL.

However, without seeing the design of the filters in the cabinets it will be hard to say definitely.

So whilst applying informed logic to your answer, mine was based on experience.

Hence in the Splitters, it is VHF/UHF Reject filters in the Phone leg, whilst the ADSL leg is "straight-through", ie no filtering.

As ADLS and VDSL both operate in the short wave and AM bands the split between pass and rejection is a lot lower than that.

See the relevant BT SIN, the specs are all in the public domain:

The insertion loss when the attached telephone is off hook between 100 Ohm source
and loads at Line and Telephony Ports should be greater than 29 dB at 32 kHz and
greater than 55 dB from 200 kHz to 1.104MHz. (2.208 MHz for ADSL2+).

Here's a splitter spec too http://ww2.pulseeng.com/products/datasheets/B899.pdf

Edited by yarwell (Sun 19-May-13 11:36:16)

I'll accept it is unlikely but it could happen.

Don't forget my experience is based on nearly 8 years designing telephone switches and line cards for various environments.

In reply to a post by MHC:

I'll accept it is unlikely but it could happen.

Don't forget my experience is based on nearly 8 years designing telephone switches and line cards for various environments.

I don't even really understand how balanced pair signal transmission works

i do recall thin wire ethernet not working if the termination resistor was removed, presumably the same thing would have happened with a short rather than an open circuit.

I'll accept it is unlikely but it could happen.

I suspect we are basically saying the same thing, it *could* happen, but is hugely improbable, hence my 99.9% answer, no.

Don't forget my experience is based on nearly 8 years designing telephone switches and line cards for various environments.

You know I know that.
If you want to make some serious spons, and cheer up a lot of CSE's, make and supply a cheap line card to replace the current system X and Y stuff which is falling over as we speak. NDT, OWT and BNR faults abound, I'm sick of getting the damn things swapped out.

One of the principal designers of the X line cards worked with me for a couple of years. At times I wanted to strangle him! He currently lives in Taplow.

He currently lives in Taplow.

Close enough, he's a marked man.
I'd ring him and moan, but his voice service is most likely OOS at present !

Anyhoo, we are drifting OT here.

His phone number is: 01628 6xxxxx


Anyway ... England are all out so I had better go and finish cutting back some 40' conifers - down to 20' WITHOUT ripping out the drop wire that runs through.

Edited by MHC (Sun 19-May-13 11:49:56)

WITHOUT ripping out the drop wire that runs through.

I feel a time related charge coming on.

In reply to a post by Zarjaz:

A damaged cable will give you audible noise, from the resulting earth and battery contacts, low insulation, HR's etc.

We've even had experience of that. The line sometimes clicks during voice calls. Not enough to annoy, but enough to know it is still there. One time last year though, it became bad enough to report, with the following symptoms (pulled off the ticket to Plusnet)

- Clicks on the line during a conversation, with no adverse consequence
- Clicks on the line during a conversation that make the exchange act like a recall button was pressed, putting call on hold
- No dial tone when picking handset up to make outgoing call
- No ringing indication from the phone for an incoming call (the caller hears ringtone)
- Incoming call cannot be picked up (handset being lifted while caller hears ringtone)
- When dialtone is not being given, a white noise can sometimes be heard - a very faint, crackly sound.

The VDSL2 signal survived all of those - it stayed 80/20, with no DLM intervention, and no change to the error rates.

[It ended with no fault found, and has reverted down to just the first symptom]

I wonder if MHC is correct - that certain problems can cause adverse effects on the way the filter works ... and I wonder if it matter what the distance is between problem and filter. That is, whether a line problem near the cabinet would have more effect than one near the exchange, or vice-versa.

I wonder if the OP has unlocked his modem to look at attainable rates ? The audible noise could well be from a mullered E-side, but his drop in sync rate is going to be something D-side related, cross talk or some such.

It would be good if true, but a little late now - there will be no "before" picture to compare against.

In reply to a post by Zarjaz:

WITHOUT ripping out the drop wire that runs through.

I feel a time related charge coming on.

So far so good ... and the span is over 45m across a main road, river and garden!

Removing the Terminating Resistor, creating an Open Circuit, would result in a Standing Wave being produced by reflection from that open end.

The Voltage doubles; and if you follow the Standing Wave back towards the Source, the voltage drops to Zero at one-quarter wavelength back, rising to a similar but opposite phase peak, half-a-wavelength back, passing through zero at 3/4s wavelength and rising again to being in phase at one wavelength back.

Similar occurs with a Short Circuit in place of the Terminating Resistor/Load; but with the zero-volts point being at the short, rising to twice the supply voltage, at 1/4 wavelength back etc.

As you surmise, the effects on operations is similar with a Short Circuit - the Ethernet will not work.

I taught this many times; but had not seen it in practice until about 1992, when I obtained a TDR (Time Domain Reflectometer) for testing Thin and Thick Ethernet installations, amongst others.

Connecting an Oscilloscope to the TDR, I was able to demonstrate all three conditions, Properly Terminated, Open Circuit and Short Circuit, to my colleagues, including the differing phasing of the Short and Open circuits.

The TDR would also measure the time delay for reflections, converting these to a "Distance to Fault" which was accurate to about 1 foot/30 centimeters. Close enough for fault finding.

The TDR could also be used on other cable types, such as mains, twisted pair etc.

A simple description of the TDR would be a"Pocket Radar Set"; and it was about the same size as a typical digital voltmeter.

---------------------------------------

Regarding the Doubling of the Voltage, if you have done any electrical loading work, the maximum power in the load occurs when it matches the impedance/resistance of the Source etc; thus the observed voltage in a simple unregulated DC circuit will show the same voltage for the Source and Load.

Remove the load and the voltage at the vacant terminals will double.

Correspondingly if those terminals are now shorted, the doubled voltage will appear in the source, leading to high temperatures etc.

Another way of explaining it is to assume that the DC voltage is an extremely low frequency AC; and that the wavelengths involved are extremely large, particularly compared to the usual DC Load demonstration circuits.

Thanks Yarwell.

All done, all finished. And no need for your services even at that discount rate you would have charged me!

My office now has natural light again!

Better buy some new blinds then.