Print

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

[kitz]

Cheers BS...  I shall try get some answers tomorrow then from PN...  and see what this is all about.

[Chrysalis]

yeah engineers ring from mobiles, I have had 3 contact me this past 12 months, all mobile numbers.

[Black Sheep]

I see you have got an engineer with you. Fingers crossed. 

[ColinS]

I see you have got an engineer with you.

You work for the NSA and I claim my £10 

[Black Sheep]

Ha ha ...... kind of, but only when the Moon is in Sagitarius, and death walks these forums !!! Apart from that, here's your tenner, pal. 

[kitz]

Grrr just typed a long post then accidentally lost it by pressing backspace in the wrong window

[Chrysalis]

so you retyping?

[kitz]

I see you have got an engineer with you. Fingers crossed. 

I did indeed, but adsl has been off for most of the day.

1) Yesterdays issue was a failure on BTs part to contact me when they should have...  from what I can piece together via PN and todays engineer:-

Engineer allocated my fault was stuck on another job and knew he wouldnt be able to make my appt so he contacted where-ever so my job could be re-allocated back into the pool for someone else to pick up.   Unfortunately no-one else was available to pick up my fault.  At this point someone from BT is supposed to contact me to rebook the appt.   Someone from BT tried to call but only let it ring twice and marked as 'EU unavailable'. 

PN are to give feedback to BT about the system and that someone should have tried harder than 2 rings to contact me.   

 However it was allocated to todays pool where the same engineer that was scheduled from yesterday picked it up this morning.

---
2)  Short version

Bit shocked to answer the door the morning to see a BToR..  anyhow he attempted to do some tests..  messed around with his JDSU several times.. muttered something about it failing on practically everything..  and says he's off to do some tests at the cab.   Comes back and says hes going to replace everything.   Goes away and comes back later this afternoon. New pair, new d-side and re-crimped every joint between my home and the cab.


----

Although he's given me the best available spare pair, in some respects its perhaps not quite as good for DSL as my old one... but tbh I am so sick and tired of this now that stability and a voice line that works has become my priority.   Id rather have a line that is decent _all_ of the time rather than one that is all over the place.


Before:
Max attainable:103172   37187
Sync      79999   20000
SNRM      12dB   18dB

After:
Max attainable: 98148   31752
Sync      79999   20000
SNRM      11dB   12.5dB

The engineer noticed these differences..  and actually offered to try another pair - but it was getting late..  and hes given me his number so I can contact him if need be and ask for him to be re-allocated since he knows what hes done and what he hasnt.

However... I noticed this

Previous power was 14.3dBm / 6.8dBm  New is 14.3dBm/4.7dBm 

My line of thinking atm is that Power has gone down..  therefore that will affect the SNRm etc. 
Based on adsl2+ theory....  Power is under control of the DSLAM/Router... and if it needs more juice then it should ask for it. 
So for some reason the DSLAM has decided I dont need as much upstream power on the new pair which will account for the lower upstream figures... and as such I dont need to worry, because if in future I need it.. then it should give it.   Thoughts?

[kitz]

oooh   I think I forgot to add some where that during a 2 second burst period ..  I experienced over 85,000 HECs and 35,000 CRCs

Is that a record?

[Chrysalis]

kitz my thoughts are you do seem to be getting lucky with engineers, 2 out of 3 now doing work, it seems opposite to my own experiences and most other reports I read on the net where engineers refuse to do much, I wish I had an engineer who redid all joints, changed d-side etc.  Any tips on getting engineer to be cooperarive?  Although I did hav an engineer change my dropwire on last visit his boss made him undo the work.

Your upstream power going down I expect is the dslam managing crosstalk, its documented by various vdsl vendors that a way for isps to manage crosstalk is to reduce upstream power, and its logical to assume this is done on short vdsl lines as they have excess signal.  This is quite possibly controlled by DLM.

Here is my output power, notice my downstream power is lower than yours.

Output power (dBmV)   13.4   6.7

Yesterday I lost sync, my upstream snrm went down to 0 (or close to it) for a short time, when the connection came back up I had a much higher upstream snrm tho, my up attainable has gone from 27meg to 34meg, and upstream crc errors are much lower than they used to be, so I think BT did some work somewhere on my loop.  However this makes my up/down sync even more imbalanced, as a line with 34 attainable on the upstream should easily manage 80mbit and my down attainable dropped to 72 now.

I dont think you have broke any records with that error rate, but your normal error rate where you said you get 1 or 2 a day is possibly a record, thats extremely low for fast path.

Almsot forgot to say, the most important thing is if your line is stable now, if it is then the new lower attainable sync is still fine, just because thats gone down it doesnt mean it wont be more stable.

[kitz]

kitz my thoughts are you do seem to be getting lucky with engineers, 2 out of 3 now doing work,

You missed some..  Ive had about 6 or 7 visits now..  Ive totally lost track.  Ive had my share of US visits.

Your upstream power going down I expect is the dslam managing crosstalk,

Yep powercut back, did actually have a discussion with the engineer about that. 
Its not part of the DLM though aiui its more to do with negotiation between the router and DSLAM and PSD masking, the idea being that your signal strength doesnt go above 'x'.    Power can be altered on the fly and can often be seen on adsl2+ being ramped up if the line starts to struggle.  Its a proper (ITU?) standard that feature across all dsl DSLAMS/MSANs.
Set masks are defined based on line length with various profiles for standard and max power. With adsl2+ these PSD masks should be exactly the same regardless of SP MSAN/DSLAM.  Its FTTC Im unsure about and know that there's possibly some additional parameters thrown in such as total line length & d-side length.

Any tips on getting engineer to be cooperarive?

Well I always offer coffee/tea/biccies... today didnt have any bacon in or would have offered a bacon butty. 
I try to be polite and despite the fact that there are some area's I know about I make sure I dont come across as condescending, even in the case when I knew one of the engineers was point blank wrong about something.   If I say something I usually try to balance it with something like ..  youre the expert with the wires and I wont pretend to know that side of things better than you.  I dont bluff either and if Im not sure about something Im happy to say so, this usually triggers a better conversation because they usually then like to tell you about their PoV.

Bear in mind though that sometimes they cant do what is needed. This may be because the assigned engineers doesn't have the skills or it could be that the fault isnt considered serious enough to warrant it.   
I guess in my case with voice now being badly affected they had to make more of an effort to track down the fault today.

dont think you have broke any records with that error rate,

Damn...   oh well its a personal best for me then.  
Never seen more than a couple of CRC per day on my line..mostly 0 for weeks.  Before the line fault I may rack up an odd few during a close thunderstorm but thats it.
I cant recall ever having seen that many CRCs/HECs rack up so quickly in all the linestats that I can recall ever having seen. Some lines may accumulate them over time...  but I thought it was rather a lot for a 2 second burst & so did the engineer who said it went waaaaaay over their fault threshold :/

[ColinS]

I cant recall ever having seen that many CRCs/HECs rack up so quickly in all the linestats that I can recall ever having seen.

Need I say anything?

[Black Sheep]

Graphs aren't 'My thang', but isn't that showing FEC, as opposed to Kitz comment about CRC/HEC ??

[Chrysalis]

if its the old graph script then FEC=CRC due to the web gui bug.

Personally I have seen 200k CRC a second before.

[ColinS]

Graphs aren't 'My thang', but isn't that showing FEC, as opposed to Kitz comment about CRC/HEC ??

Yes it is, and yes, they were all recovered. The technology works.  The numbers are astonishing though, although this only happens very rarely and decreasingly so lately.  ;D  I'm not worried about it.