Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ColinS]

While researching something else entirely different  (cf Monty Python), I was reminded of this observation, which I thought we all might recognise from our various experiences with changes in BB speed estimates, fault-finding and the like, irrespective of CSP/ISP

After the Challenger Shuttle disaster the term "Normalisation of Deviance" was coined with respect to failure of the "O" Rings which caused the catastrophic failure. The deviation from acceptable had become so routine that those involved subconsciously thought this was "normal".

[kitz]

Thanks for the comments and support, guess I was just a bit weary yesterday.   PN have booked another appt for me, its just the thought of going through it all again.   Line has been generally been ok the past few days - except for when the phone rings and is in use.  It does appear to get worse when the weather is hot & dry.

----

Using the phone will wipe out all my bitloading up to tone 96 (414 kHz) and will at least halve the bit loading in all frequencies in up to tone 1026.
All of my bitswap activity is in those lower tones, which is highly unusual, as normally you'd expect bitswap to occur at the higher frequencies

[burakkucat]

You must not 'throw in the towel', for PlusNet are on your side.   

Just one thing to consider. If PlusNet are also your telephony provider (I forget your precise details), are they referring the fault back to Openreach as an intermittent voice fault? (Crackly line, etc.) If not, why not?

You should also stress to PlusNet -- so that they make it clear to Openreach -- that the fault becomes so bad at times that you have to resort to using your mobile telephone. In other words the telephony service is not fit for purpose and must be repaired. 

I would not be surprised that if a network engineer was booked to attend and given instructions to re-crimp each joint in the pair between the PCP and your home the fault would then be fixed. 

[kitz]

>> If PlusNet are also your telephony provider

Just had my 'Goodbye' letter from BT this week.  I think PN wait a few weeks after the dsl has gone over to FTTC before taking over the phone line.

>> that the fault becomes so bad at times that you have to resort to using your mobile telephone.

I told the OR guy last here that at times its been so bad that either I or the caller have had to ring back again.   Resorting to the mobile is lazyness on my part to save me the possibility of making 2 calls...  and my daughter calls me on my mobile for same reason.  Plusnet now also ring me on my mobile because they too have experienced a bad call on the landline and couldnt hear me.

>> to re-crimp each joint in the pair between the PCP and your home

This is something I have wondered about since the engineer definitely disturbed something just by testing from the cab.   I dont know enough about BT testing to know though why simply him testing at the cab would have a permanent affect on my linestats.   

[burakkucat]

This is something I have wondered about since the engineer definitely disturbed something just by testing from the cab.   I dont know enough about BT testing to know though why simply him testing at the cab would have a permanent affect on my linestats.

   Hmm . . . Assuming the PCP, through which your pair is connected, is quite normal then there would have been a pair of crimps joining the D-side and E-side wires of your pair together. When a NGA GEA FTTC VDSL2 service is provisioned on the pair, those two crimps are removed, the D- & E-sides are separated from each other and are then crimped to the respective tie-cables that link the PCP to the FTTC. So two crimps are replaced with four crimps.    I presume that any invasive testing would probably result in the removal of the two crimps that connect your D-side pair to the D-side tie-cable pair (link to the FTTC). If the subsequent re-made joint(s) were somewhat defective . . . 

Perhaps Black Sheep could indicate what would be the most likely method used, when testing from the PCP, please?

[kitz]

Eurika!

I think today's engineer has found the problem.    Engineer this morning understood and could see the large fluctuations in my upstream whilst the phone was in use...  and said he was off to change the d-side.   A few mins later he called me to 'come look at this' as he noticed that the d-side to the BT66 was fine.    He showed me the tests on his JDSU and sure enough the line was rock steady even with the phone in use and phone ringing.   

So he comes back inside, and plugs his JDSU again from the test socket...  line starts fluctuating and upstream dipping, along with the max sync dropping.  So its not the fttc i-plate.    Removes the NTE completely and tests from the bare wire.    Line is fine.
Repeats the test, puts the NTE back on, trys the test socket..  line plays up. 
Gets a brand new socket, installs it...  and line is perfectly fine.

Ive no idea where the NTE came from, but according to todays engineer its not a new one, as he can tell by the badging.  It's possibly the same one thats been moved from the old NTE location.   I think the engineer was a bit surprised as he said its not an obvious failure point and not something they normally look for.  I repeat this is not the face plate, nor the i-plate that failed, but the back part of the NTE

For the first time in several weeks my SNRm doesn't budge when the phone is in use nor when the phone rings.  I really really hope that this is what it was.   

[HPsauce]

Fingers crossed.  8)
Don't you just love it when someone actually applies logic and discovers the actual cause of a fault, especially when it's rather out of the ordinary. 

[roseway]

Here's hoping!

[ColinS]

Go on, admit it Kitz - you got Black_Sheep to come round personally.

No, all due credit and admiration for the engineer who took the time and trouble to identify and fix your fault.
 :clap2:

You've got to send in some sort of commendation to his boss in BT.  It's so different, sadly, from most of the visit reports we hear, that it is genuinely astonishing - in the very best sense of the word.

[kitz]

Today's engineer said that its usually the downstream that indicates that there is a problem with the line.  He even admitted that when he saw my upstream fluctuating all over the place he wasnt sure if it was a normal feature.. just because its something that they would never normally look at.

However, by using his JSDU to monitor what the upstream SNRm was doing that was what allowed him to pin point where the fault was actually coming from. 

It was only after being told my observations about the upstream, that he decided that he was going to use this info to attempt to trace the fault and work backwards from the BT66.  TBH Im glad he did..  because if this is what the problem has been all along then changing the d-side would have done nada.


-----

It only dawned on me once he'd gone that the [bad] NTE is either the one most likely installed when the house was built and the one that the Quinn guy moved, or it came from the Quinns guys collection of old kit.  It definitely cant have been replaced when the 1st BToR engineer re-wired because its not new enough.   If it is the original, then it will have been backing right on to where all the oxidisation was in the photo showed earlier in my thread.

[kitz]

>>> Go on, admit it Kitz - you got Black_Sheep to come round personally

haha... nope, but I do believe he was I believe an SFI guy, who is given a bit more time to look into things... and as such he took on-board what I was saying about the upstream SNRm.   Usual  BToR engineers would not look at SNRm. 

Credit too has to be given to the authors of monitoring programs such as DSLstats and HG612_modem_stats, because those tools have been invaluable to point out what was going on with my line.   Without them, I would just have a line that would drop for no reason and not know why.   So a personal thank you to Eric for not only writing DSLstats, but also taking on board the suggestions Ive chucked at him whilst this fault has occurred..   I also must not forget Bald Eagles scripts because they and everyone else involved in graphing & hacking the HG612  have helped tremendously too. 

Anyway as I write this, I glance sideways to the other monitor as see two totally flat lines graphing SNRm for the past 2 hours.   Lets hope thats it now.

[ColinS]

If it is the original, then it will have been backing right on to where all the oxidisation was in the photo showed earlier in my thread.

Sounds quite likely.  Hope he took it away, takes it apart & shows it round the linesman's room then.  But again, all credit to him for listening to the customer, and being determined to follow through on that, despite what they might 'normally' do. 

[kitz]

>>> Hope he took it away,

He did indeed.

hahaha that would be fun if they tested it on a more normal line.   IF that is what been causing a normally rock steady line with 18dB of margin to drop out..  christ knows what it would do to an average line. 


---
BTW Ive never recovered the 2dB of SNR since last weeks testing from the cab...   
However in the grand scheme of things, I will accept that as 'one of those things'...  Im not going to go bitching about it as long as I have my line stability back which is the main thing as is being able to use the phone and broadband at the same time

[burakkucat]

Well that is some good news! 

My thoughts on the old NTE5/A. If it was one of the very earliest design with the gas discharge over-voltage / surge-clamping device fitted to it (just like the one in The Cattery), then it could be that the problem was a result of that simple device going faulty.

In a different thread, I posted that they glow a nice violet colour when ignited by an application of 500V DC.

[kitz]

Thank you b*cat ... It had a white casing not green, so I dont think it was quite as old as that one.  TBH I didnt look too closely at the terminals because the engineer opened it up and looked at the circuit board suggesting that it was likely something on the board.

If this was the original NTE (which I now suspect it likely was)...  and remembering the state of the BT66 which it directly backed on to, then it wouldnt surprise me if the circuit board had been exposed to the same damp conditions which had totally corroded the terminals inside the BT66.