Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Author Topic: Help With Hard to Find Fault - Possible High Resistance  (Read 466 times)

siofjofj

  • Member
  • **
  • Posts: 54
Help With Hard to Find Fault - Possible High Resistance
« on: September 15, 2020, 04:09:40 PM »

Hi all,

This topic is related to the same fault that we were discussing here https://forum.kitz.co.uk/index.php/topic,21455.0.html, however a significant amount of time has elapsed and I now believe that the initial assumption I made in that thread (MSAN ports going bad) is incorrect, so I thought a new thread would be worthwhile for clarity. Sorry for the very long post, but this story is somewhat complicated.

In a nutshell, the problem I am having is repeated loss of ADSL sync, occurring on an intermittent basis. We can go for a few days with an absolutely solid connection and not a single CRC error, however this can then change to a completely unusable connection for several hours with the modem staying in sync for only seconds before re-syncing over and over again. On average, I'd say the connection is unusable for about 10% of each day, and extremely frustrating outside of that with random disconnections occurring.

When the connection is playing up, the symptoms are as follows, as tested with a known good modem and corded telephone connected to the test socket using a known good microfilter:

  • Large numbers of CRC errors and loss of sync occurring at random, but frequent, intervals (see attached image crc.png for typical example)
  • Loss of sync occurring when the telephone is taken 'off-hook'
  • Large number of CRCs occurring when the telephone is ringing (often followed by loss of sync)
  • Significant oscillation of the SNRM (see attached snrm2.png for typical example)
  • High pitched white noise clearly audible on the phone line while the modem is in sync, disappearing when sync is lost before restarting.
  • Occasional crackling audible on the line, whether or not the modem is in sync

This problem has now been going on for roughly 2.5 years and, with me now having to work from home which involves numerous video calls, is becoming unbearable. In this time we have had visits from roughly 10 Openreach engineers, none of whom have managed to fix it. The results of the tests performed by the Openreach engineers on every visit are as follows:

  • Exchange-based 'Fast-Test': Often passes, sometimes fails with 'CIDT Fault'
  • House-based 'Pair-Quality' test. Always passes
As a result of this, a large number of Openreach visits end up being 'Right When Tested' so nothing gets done. Sometimes, the engineer performs a 'DSL Close-Out Test', using their tester as a modem and monitoring for CRCs and loss of sync. If the line is having a good period at that point, they see nothing, so again we have 'Right When Tested'. If the line is acting up at that point, they see the errors, particularly when the telephone is picked up or if I sneakily ring the phone from my mobile during the test.

If the errors were seen during the test, on every case they have gone away and done something, usually swapping between pairs at various different junction boxes until no more errors are seen during the DSL close out test. On a few occasions, the port on TalkTalk's MSAN was changed (hence the other thread). When they come back to the house, the DSL Close-Out Test passes with no errors and they leave. Without fail the fault then reoccurs some time (even minutes later in same cases) after they have left.

Lately, we have changed ISP to NowTV. Partially this was because of the suspicion of a bad line card on TalkTalk's MSAN (see other thread) so, since Sky have their own LLU equipment, we thought this would fix the problem. It didn't, the problem is just as bad as before.

Two weeks ago, we had an Openreach engineer come out and he saw the huge number of CRCs on his tester. He then went to the joint box in the pavement on the opposite side of the road to the house and did the same test there, finding no errors. He then swapped us to the green/black pair in the cable from the joint box to the house and tested again at the house. This time, he got even more CRCs than when testing at the house the first time. He therefore put us back on the orange/white pair and put in a job for the replacement of the 'lead-in' (the cable between the joint box and our house). The house is 1970s build with an armoured telephone cable buried directly in the ground from the joint box on the other side of the road, up the street a bit.This means this job is very difficult and expensive as it requires digging up our front garden, the pavement outside the house down the street and then the road to get across to the joint box on the other side.

Last week, a civils team together with a PSTN engineer came along. Apparently the job had been changed by management from 'change the lead-in' to 'find the fault on the cable run, then dig there and repair it'. Of course, no-one knows where on the lead-in the fault was, so they guessed at under the pavement outside my neighbour's house as they thought it might be the buried joint there. They dug a hole in the pavement, but found no telephone cables in the hole so filled it in again and left.

A few days later, we had another broadband engineer visit who put in for replacement of the lead-in again (though he was unable to reproduce the fault as the line was fine at that time, he did believe the previous engineer who said that there is one). Yesterday, another civils team turned up. They dug up the front lawn and put in ducting for a new lead-in between the house and pavement. So far so good. However, they ran out of time so had to come back today to dig up the pavement and road to get to the joint box. This morning they arrived with a broadband engineer who tested the line and couldn't find the fault (line having one of its good periods), so he told them to cancel the lead-in change, leaving the newly installed duct in the front lawn empty. Instead, they dug in front of next door's house again and the engineer spliced my lead-in onto the spare pair on next door's lead in (this effectively replaces only 5m of the 30m lead in). As soon as the engineer left, but while the civils team were still there, the disconnections restarted. I showed one member of the civils team the disconnections and the noise on the phone, and he agreed that it is still broken, so he tried to call the engineer that just left. We were told to simply report the fault again via our CP.

This is a complete nightmare, since going back to reporting the fault via NowTV leaves us with a 50/50 chance they will find nothing again, so we will have to keep trying appointment after appointment until an engineer happens to come when the problem is apparent. Then the issue is that if the engineer can't fix the problem that day, the next engineer does the tests again and cancels the job if the fault can't be found the next day.

My questions to the good people at kitz are as follows:
1) Does anyone have any ideas about a better way to approach Openreach or NowTV about this, because right now I am getting nowhere.
2) Does anyone have any other ideas about what the actual problem with the line might be? It may be that the fault simply disappeared (in the way it does) when the engineer tested from the joint box, so perhaps the lead-in is not the problem after all.
« Last Edit: September 15, 2020, 04:15:38 PM by siofjofj »
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 30477
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #1 on: September 15, 2020, 05:05:37 PM »

All evidence that I have examined points to one (or more) joint(s) showing HR or semi-conductive tendencies. As for the location of the defect, I cannot say . . .  :no:

Attaching a TDR to the end of the lead-in at your home and making a series of measurements should give a good preliminary location. Going to the PCP and making a second series of TDR measurements, then cross-referencing both sets of results should narrow down the target area.

Next perform a sequence of classic resistive fault location (RFL) tests. Firstly from your end of the lead-in and then at the PCP.

Cross-referencing both sets TDR and RFL data will pin-point the faulty plant.

What can you do? My suggestion is that you compose a concise e-mail message, enclosing all details of the relevant facts, dates & times, and send it to clive.selley@openreach.co.uk, requesting a higher level intervention.

[Edited to correct a typo.]
« Last Edit: September 15, 2020, 06:59:40 PM by burakkucat »
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

siofjofj

  • Member
  • **
  • Posts: 54
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #2 on: September 15, 2020, 05:28:34 PM »

Thanks burakkucat. I saw the suggestion of emailing Clive Selley in another thread, so was indeed wondering if any kitizens thought my fault was 'worthy' of such an intervention.

Do you think it would be worth including line stats with such an email? I have attempted to show some of the visiting engineers the stats in case the information would be useful to them, but none ever seemed interested.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 30477
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #3 on: September 15, 2020, 06:58:23 PM »

I saw the suggestion of emailing Clive Selley in another thread, so was indeed wondering if any kitizens thought my fault was 'worthy' of such an intervention.

It certainly is a worthy case.

Quote
Do you think it would be worth including line stats with such an email?

Just mention that the information is available and can be supplied, if necessary.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Black Sheep

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5118
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #4 on: September 15, 2020, 07:24:16 PM »

I totally concur with B*Cat .... this is the 'High Level Complaints' team, bread & butter type jobs.

An absolutely certain high resistance fault (going from your description) ... and shouldn't be 'that' hard to find, TBH ?. Luckily, the HLC team will generally ensure the right man for the task attends site.

Go for it.
Logged

siofjofj

  • Member
  • **
  • Posts: 54
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #5 on: September 15, 2020, 08:32:49 PM »

Thanks guys. I'll get on it.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 30477
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #6 on: September 15, 2020, 10:14:35 PM »

. . . the right man for the task attends site.

Preferably with a Tester 301C and an Ohmmeter 18C:D  ;)
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

Black Sheep

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5118
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #7 on: September 16, 2020, 10:58:39 AM »

Preferably with a Tester 301C and an Ohmmeter 18C:D  ;)

Ha ha ..... proper meters !!  ;) ;D
Logged

siofjofj

  • Member
  • **
  • Posts: 54
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #8 on: September 20, 2020, 05:23:40 PM »

Well, that looks to have been very successful!

On Friday, a few days after sending the email, the Openreach 'patch lead' (i.e. the leader for the area) turned up at the house. She had read all the notes from the previous engineer visits and asked us for our description of the problem. She was very apologetic for how difficult it had been to get this fixed, and said that one of their best engineers would be sent to the job tomorrow (Saturday!) and would be able to stay on it until we were happy it was fixed (or alternatively until a plan for fixing it could be established if it was not possible to fix it that day).

On Saturday, sure enough the engineer arrives. We discuss the fault and what has already been done by other engineers, and off he goes to investigate. He ultimately concluded, after spending all morning looking, that the fault was not on our lead-in but somewhere else in the network. He found a spare stopped line in the joint box opposite my house, so swapped us onto that. The connection has been solid for 36 hours now and the phone is completely silent. I'm currently using NowTV's supplied router so can't see if we are getting CRCs when the phone rings (hardly any stats visible on this), but I will swap back to my own when NowTV are happy it is stable so I'll be able to check this later. The engineer even said he would return on Sunday to check I was happy, but I texted him this morning to say he didn't need to come since all was well. He said I can contact him directly if the issue reoccurs.

So, thank you for the advice burakkucat and Black Sheep, it worked out really well!

I feel very sorry for the engineer that diagnosed the lead-in and caused (apparently) 3000 worth of digging to be done. It looks like he got caught out by the fault disappearing (as it does) when he got to the joint box and then reappearing when he got back to the house.

I also feel sorry for the civils team who, while digging up the front garden made a shocking (thankfully not literally) discovery. During the dig, they found a duct with 'G.P.O.' written on it. This was a surprise because, as seen at the joint box and at the house, the lead-in appeared to be directly buried in the ground. However it was (seemingly) good news since this would mean the entire route would not need digging up. They therefore broke into the duct to allow pulling in a new cable, only to find a thick cable marked 'POWER' inside it. The house builders seem to have used G.P.O. branded ducting for the electricity supply which, in my opinion and the opinion of the civils guys, is utterly ridiculous! This duct was then swiftly repaired.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 30477
  • Over the Rainbow Bridge
    • The ELRepo Project
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #9 on: September 20, 2020, 05:48:17 PM »

A good result. Thank you for sharing the news.  :)

Just in case you are not aware of its existence, there is a "Thank an Engineer" page that you now might like to use?
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

siofjofj

  • Member
  • **
  • Posts: 54
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #10 on: September 20, 2020, 05:56:36 PM »

Indeed I would. Thanks for the reminder!
Logged

Black Sheep

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5118
Re: Help With Hard to Find Fault - Possible High Resistance
« Reply #11 on: September 21, 2020, 09:06:05 AM »

To reiterate what B*Cat has said, thank you for the feedback.  :)
Logged
 

anything