Confused!!!

[#1]

Added those margin graphs to the ticket, see what they make of it.
As I was typing this, the router just lost the connection to plusnet, but didnt lose sync.  Coincidence or something more sinister

[#2]

Have you noticed whether there is any particular behaviour of the upstream NM when you find the line is hissy/crackling? I know you've said that if the line drops the US sync is usually low and the US SNRM goes flat.
Whichever, next time you hear the line hissing/crackling, just ring your phone from your mobile, let it ring a few times, watch what's happening to the SNRM, up as well as down, don't answer it. Cut the call, and then see whether the line is still hissing and crackling and what the SNRM then does.

[#3]

@anotherone... Firstly, i would like to say i really appreciate the interest you are taking with this fault... As you can appreciate we both feel there is an intermittant line fault... Wish outofreach felt the same. Will try.your suggestion, and let you know...

[#4]

OK.  Still no resyncs with the standard quality filter in the test socket....
Checked line with a quiet line test. There was a bit of hissing.  Phoned house phone with mobile, 3 times (see margins plots).  After the first 2 times there was still some hissing on the line.  After the 3rd time, the line is completely quiet, no hissing whatsoever.
Seems to me to be marked drops in SNRM, indicating noise on the line, but there are more knowlegable guys out there than me...  If BT do come out, would the engineer be interested in the margin plots??  When this saga first started again, I was getting sync drops with incoming and outgoing calls.  After the engineer vists at least that has stopped now....
unsure why the stats graph tells me upstream interleaving is off......

[#5]

Ok. It seems to me that proves there is an HR joint somewhere or a problem with the splitter in the exchange. What we can't show is if it's an HR joint, whether it's on the line or at the exchange or even at your end.
As you know, ringing current can clean up a joint, usually only for a shortish period (relatively).
Yes the drops in SNRM show that the ringing current is interfering which it can only do if there is a HR joint or faulty exchange splitter.
Notice now that since the problem had re-occurred the US margin had been floating around 22/23dB even dropping to around 18dB when the US sync was low and now that the noise has disappeared the margin has improved to 29dB. The US attenuation is also the lowest in all the stats you've posted in this thread now 23.6dB (previous best was 23.8dB, worst was 24.1dB). Yes an engineer may be interested in the plots, but if he's POTS it will depend on how clued up he is.
Next time the problem shows itself, US SNRM down at around 22dB or worse, try switching off anything your end that radiates RF - cordless handsets and basestation and all mobiles etc. except the one you are going to use to ring the landline. Go as far away as possible from your line - bottom of the garden sort of thing - ring the landline  and see what happens to the US NM trace. If it's still all over the place and eventually goes to ~29dB after a few rings that would suggest it's not your end. If however, it stays flat even if it jumps back to about 29dB then I'd be looking for a dodgy connection say the NTE5 (or any other joints at the house on the drop wire) or modem/router lead - the only common factors at present (apart from the modem/router but you've changed it before with no change in the problem right?).
Can't think of anything else at the moment.

[#6]

Quote from: Anotherone

Ok. It seems to me that proves there is an HR joint somewhere or a problem with the splitter in the exchange. What we can't show is if it's an HR joint, whether it's on the line or at the exchange or even at your end. As you know, ringing current can clean up a joint, usually only for a shortish period (relatively).. Next time the problem shows itself, US SNRM down at around 22dB or worse, try switching off anything your end that radiates RF - cordless handsets and basestation and all mobiles etc. except the one you are going to use to ring the landline. Go as far away as possible from your line - bottom of the garden sort of thing - ring the landline  and see what happens to the US NM trace. If it's still all over the place and eventually goes to ~29dB after a few rings that would suggest it's not your end. If however, it stays flat even if it jumps back to about 29dB then I'd be looking for a dodgy connection say the NTE5 (or any other joints at the house on the drop wire) or modem/router lead - the only common factors at present (apart from the modem/router but you've changed it before with no change in the problem right?).

already used a new router (well not new, but never been used, the parents Thompson 585V7)  Still the same.  Only thing I dont think I havnt changed is the Router lead.  Its a quality screened one, from Broadband buyer, had it a number of years now.  Will put on a standard lead, and then go through your checks, and switch off everything I can, although not sure what may radiate RF to the phone, we dont even have a radio in the house.....  I have a couple of joints in the house.  1 at the NTE5, at the BT demarcation point, and another on the windowsill where the dropwire comes in.  Its a little junction box, Thick black lead in, standard CW1308 out to NTE.  This CW1308, does pass next to the mass of cables behind the TV, so I will also give this a good tidy up as well, move all power leads away from phone lead etc.... I suppose I could be a bit naughty and check these joints out.  Both BT engineers didnt bother.
If needed, do you think i should be getting a POTS or an SFI engineer out, or leave this to Plusnet.....
Whats baffling me is why the XTE made things unstable, and a standard filter has stabalised the line, or may be just coincidence, or the will of the gods at BTO

[#7]

It could be just coincidence. The active parts of the XTE could complicate what you see so it's best out of the picture for the moment.
I'd check those joints yourself for starters, if you find an issue you can't quietly deal with then it depends on what else is going on.
At this point I don't think an SFI is warranted, as it is looking like a POTS issue dependent on the location.

[#8]

Thanks....  I will check the back of the NTE, and the small junction box connections for anything obvious.  I suspect if it was anything there, it would be obvious.  Will also trace the route of the cable in the house, to make doubly sure that its not damaged, although I am nigh on sure its fine....
Plusnet appeared to have washed their hands of this.  Had a response to my ticket as follows:---
[quoteDear Mr Hesketh,
Thank you for your patience while we investigate your problem. Please accept my apologies for any inconvenience caused.
Your issue appears to be resolved as your connection is now stable after a period of disconnections. Please leave your equipment connected as it is at the moment and continue to monitor this and let us know if you experience any further issues.
You can get back in touch with us online at http://contactus.plus.net or by phone on 0800 432 0200 or from a mobile on 0345 140 0200.
Kind regards,
]
No indication they have even looked at the noise margin graphing I have attached......  Yes the line has stabalised, NO THE ISSUE IS NOT RESOLVED......  Can a member of the Digi care team take this up please.  I really feel that Openreach need to rule out any joints on the D side for starters, do line pair quality tests, and then take it from there.

[#9]

Clearly no-one with any technical knowledge has looked at it 
Did your ticket clearly state the problem was intermittent? If so, was it the same agent that has just responded, well whichever if you clearly stated it was intermittent my ticket reply would be extremely blunt to point of almost being very rude about agents reading ticket history carefully, or even at all.

[#10]

Hi Penfold,
I'll get this picked back up for you and investigated, I'll give them the history of your issue to make them aware.
It looks like you got the response because over the past 4 days your line has been stable.
Our faults team will be back in-touch.

[#11]

Thanks Chris
Well after that little flurry of stability, according to the router logs, had a couple of resyncs today, together with what appears to be difficulty in getting a PPP connection as well....
Resynced at a lower upstream (404) and a lower downstream as well
ADSL Link Downstream Upstream
Connection Speed 6816 kbps 403 kbps
Line Attenuation 46.5 db 23.8 db
Noise Margin 5.2 db 1.3 db
As you can see upstream margin is terrible, and upstream attenuation going up again, all pointing to an HR fault.  Line currently noisy, and routerstats graphs attached, with a phone call from bottom of garden      Everything is switched off thats anywhere near line (LCD TV, Bluray, DVD etc  I also mean switched off, not just on standby)
The graphing quite clearly states to me that there is an HR fault on the line somewhere.  It has been very bad weather today in blackburn.... 

[#12]

Did you look at the connections that the eng didn't bother to?

[#13]

Hi Anotherone....  Not had time today to be honest, working from 6am till around 7pm tomorrow as well.  As soon as I get chance I will check the wiring out, but not keen on climbing the pole     Does my hypothesis sound OK to you, with regards to the HR fault.  If I am having difficulty in getting a connection to PlusNet as well, (Router logs suggest this with lots of LCP Down, LCP allowed to come up, LCP Down etc etc, before I get a CHAP Authentication sucess), could this all be a faulty line card at the exchange?

[#14]

Well I suppose it could be, BUT if you have a shed load of US noise at the time then US will be readily corrupted and so getting CHAP could be troublesome until the noise dies off. A plot from RS at the time might throw more light on that.

[#15]

Well an engineer is coming out monday aft... Lets see what they find, i will however insist at the very least they remake the dropwire - pole connection. Lets see what they make of the snrm graphing as well... Had quite a few resyncs today, according to the logs...