Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: Black Sheep on November 17, 2013, 12:27:35 PM
-
As I'm waiting around to be taken to the boozer by my missus, I thought I'd do a quick WHOOSH test against my own VDSL circuit.
My DS is 65.7Meg, and my US is 19Meg. My service profile is ... 40M-80M Downstream, Interleaving Low - 10M-20M Upstream, Interleaving Off and has only changed the once, from its initial 128Kbps-80Mbps profile on start-up of the product. I was getting the full 80/20 Meg speeds on start-up.
I was the 2nd EU to go 'Live' on my Cabinet, my circuit is 0.5mm Copper all the way from the PCP, and believe me when I say it tests as good as any line could wish to be. There are now 36 EU's with VDSL service on this PCP
So, I can only deduce that 'Crosstalk' is alive and well, and is the culprit behind my 15Meg deficit ?? I thought it worth mentioning as quite a few contributors don't believe that Xtalk is as big an issue as it is, and think engineers 'use it' as an excuse to fob them off with. Roll on vectoring, eh ?? :)
----
Edited by Admin
(added crosstalk to topic title)
-
15Mb seems an awful lot to lose for crosstalk. Im pretty sure BE would agree with this statement.
Im still connecting at 80/20, but have noticed my max attainable that at first was @ 103.5Mb is now 88.5Mb
Upstream was circa 35.5Mb and is now 30Mb. So thats a 'loss' of 15Mb for me too.
-
For some of us, who have an obsessive curiosity, it would be interesting to know just how low your connection's DS Interleaving depth is as mentioned here:-
"40M-80M Downstream, Interleaving Low - 10M-20M Upstream, Interleaving Off"
I am guessing that at a speed of 65.7Meg you probably haven't noticed much in the way of a deteriorated service?
If you recall, after my connection issues were finally fixed, I was able to achieve sync speeds of around 30 Mbps for quite a few months before it gradually started to decline to its current 20 Mbps.
This decline was put down to crosstalk as all physical line tests resulted in LTOK.
Crosstalk as more users have been connected is quite possibly the real reason as I did see a stepped deterioration starting early January 2013.
I can actually see/feel this deterioration, especially when working from home over a VPN connection to my employer's server(s) & my general internet 'experience' has also deteriorated.
At the time I eventually reported the speed decline (late April/early May), my DS speed estimate was still 30 Mbps.
i.e. my sustained sync speed was more than 25% below estimated speed, which warranted a number of engineer visits to investigate.
Apparently, my cabinet in now almost 'full', therefore various visiting engineers assumed the deterioration must be due to increased crosstalk.
Within a few days of reporting the speed deterioration, the response was to lower my DS speed estimate to 22 Mbps, therefore my connection is supposedly now performing "within acceptable limits".
I did think that was a bit of a 'cop out' as there was no conclusive evidence that crosstalk had indeed caused the decline.
Now then, I notice the format of the BT BROADBAND AVAILABILITY CHECKER has changed this morning:-
https://www.dslchecker.bt.com/
It now shows estimates for FTTC Range A (Clean) & FTTC Range B (Impacted).
Unfortunately, there is no explanation of what 'Impacted' means.
Do you know if 'Impacted' is some sort of reference to connections impacted by crosstalk (or something else)?
For my connection, the estimated speeds are actually the same for Clean & Impacted.
See the attached screenshots, firstlky for my estimates before today & secondly today's estimates.
-
If you check now BE they have added
"For FTTC Ranges A and B, the term "Clean" relates to a line which is free from any wiring issues (e.g. Bridge Taps) and/or Copper line conditions, and the term "Impacted" relates to a line which may have wiring issues (e.g. Bridge Taps) and/or Copper line conditions."
This new revised estimates were added last night at some point as I checked around 6pm last night.
Just to point out both my speeds are the same as well.
-
I have a question for BE1, before the decline of thoughput speeds did you notice anything untowards in your stats like big changes in SNRM's and also any high errors ?
-
If you check now BE they have added
"For FTTC Ranges A and B, the term "Clean" relates to a line which is free from any wiring issues (e.g. Bridge Taps) and/or Copper line conditions, and the term "Impacted" relates to a line which may have wiring issues (e.g. Bridge Taps) and/or Copper line conditions."
Which URL are you using?
It doesn't show that detail for my connection when using https://www.dslchecker.bt.com/
-
http://www.dslchecker.bt.com/adsl/adslchecker.welcome
-
I have a question for BE1, before the decline of thoughput speeds did you notice anything untowards in your stats like big changes in SNRM's and also any high errors ?
See the attached (320 days of stats).
It's not just throughput speeds, it's actual sync speeds & attainable rates that hve deteriorated.
Due to the scale of the graphs over 320 days it's not all that clear, but I did see lowered SNRM for a few days before each lower speed resync.
The most noticeable was from 08:28, 31/12/2012, although the connection didn't resync at a lower speed until 01:18, 05/01/2013.
-
http://www.dslchecker.bt.com/adsl/adslchecker.welcome
Thanks for that link. It does now include the explanation for my connection.
-
gonna look at your stats carefully ;)
I am having issues as you know and just did availability check and my Cab Number has changed :'( could this be my problem ?
see below.
edit never mind could of swore it said cabinet 4
-
[off topic]
If the End User wishes to migrate from their current Broadband supplier they will need to contact them in the first instance to obtain a MAC (Migrations Authorisation) Code, and then contact their new Broadband supplier to arrange for the service to be migrated.
I would have thought, given her age, that by now Beattie could construct a sensible sentence . . . ::)
If the End User wishes to migrate from their current Broadband supplier they will need to contact them in the first instance to obtain a MAC (Migrations Authorisation Code) and then contact their new Broadband supplier to arrange for the service to be migrated.
Big hints: Remove the redundant comma that follows the word "Code" and precedes the word "and". Move the closing parenthesis that follows the word "Authorisation" and precedes the word "Code" so that it follows the word "Code" and precedes the word "and".
:wall:
[/off topic]
-
[off topic]
I would have thought, given her age, that by now Beattie could construct a sensible sentence . . . ::)
I am stunned :blush: how were you able to determine the Sex of Beattie
-
Whenever someone types "Beattie" I have an instant image of Beattie Bellman (http://en.wikipedia.org/wiki/Beatrice_Bellman) & not BT the teleco.
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fcdn-static.cnet.co.uk%2Fi%2Fc%2Feblogs%2Fjasonjenkins%2Fbeattie.jpg&hash=f76277dded364a263857382d11077a9858c85b32)
-
Whenever someone types "Beattie" I have an instant image of Beattie Bellman (http://en.wikipedia.org/wiki/Beatrice_Bellman) & not BT the teleco.
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fcdn-static.cnet.co.uk%2Fi%2Fc%2Feblogs%2Fjasonjenkins%2Fbeattie.jpg&hash=f76277dded364a263857382d11077a9858c85b32)
Thats a Drag Queen for sure :D
-
Whenever someone types "Beattie" I have an instant image of Beattie Bellman (http://en.wikipedia.org/wiki/Beatrice_Bellman) & not BT the teleco.
Some good memories here (http://www.tellyads.com/tellyads_beattie.php). ;) It's just unfortunate that all the links return a "404". >:(
-
They are all here, part way down the page:-
http://www.tellyads.com/vintage_sub_index.php?alpha=B
-
Black sheep, sorry I only just found this thread.
I actually thought you told me you never see big crosstalk before :) (apologies if you never said that)
I was one of those saying I think crosstalk is a big problem at least for some lines.
so if you want my opinion is 15meg crosstalk realistic on a line that synced at 80mbit? absolutely.
My line I have estimated roughly 40mbit lost to crosstalk down from 110 to just under 70. Plus 9mbit lost on the upstream from 36mbit to 27mbit.
-
For some of us, who have an obsessive curiosity, it would be interesting to know just how low your connection's DS Interleaving depth is as mentioned here:-
<snip>
yeah Whilst your issue probably is crosstalk, I agree with you entirely that really BT should be doing full physical checks before diagnosing crosstalk. But its probably cheaper to do a pair quality test, if it says ok and then just decide its crosstalk at that point.
-
It's updated and changed (again)!
High Low High Low
FTTC Range A (Clean) 28.7 20 6.7 4.9 -- Available
FTTC Range B (Impacted) 23.1 11.2 5.9 2.9 -- Available
Clean is higher than my original estimate!
-
Same here! Whats going on, anyone know? :-\
-
It's for self-installs (rumored) which will be going live early next year!
I also just wondered what estimate the ISP will give you out of them!
BT
20Mbps* Download speed (Clean, Low)
4.9Mbps* Upload speed (Clean, Low)
Sky
20.3Mbps Download speed (Clean, Low?)
N/A Upload Speed
TalkTalk
20Mbps Download speed (Clean, Low)
4Mbps Upload speed
Mixed results!
-
Hi. Customers
On the provision of a new instalition from the Cab D side a high and low insulation test should be taken to the Master Socket. ( speed point )
This consistes of A and B lines to Earth plus loop resistance in ohms these results should be given to the customer in their contract with there ISP.
This test will not pick up slip pairs on an unbalanced cable but these will be corrected by the special faults group in my day.
All overhearing faults if were difficult or vague were passed to SFG.
Regards. Phil
-
Hi. Customers
On the provision of a new instalition from the Cab D side a high and low insulation test should be taken to the Master Socket. ( speed point )
This consistes of A and B lines to Earth plus loop resistance in ohms these results should be given to the customer in their contract with there ISP.
This test will not pick up slip pairs on an unbalanced cable but these will be corrected by the special faults group in my day.
All overhearing faults if were difficult or vague were passed to SFG.
Regards. Phil
Source of this?
I also believe this to be nothing new...
-
I might be more convinced by this checker but I used the numbers of 3 neighbors. I would guess a range of distances of 220m to 350m to the CAB. Each of them nearer the cab and find the same totally identical result to my own!! ???
-
I might be more convinced by this checker but I used the numbers of 3 neighbors. I would guess a range of distances of 220m to 350m to the CAB. Each of them nearer the cab and find the same totally identical result to my own!! ???
Estimates are up to the DP, I assume all three of you are fed via the same DP.
-
Here's what they now estimate for my number. Seeing that the old style estimates were 26.2 Mb/s and I am syncing on ADSL2+ ~7.5-8Mb/s, not that attractive. I'm 650 -850 metres from the cab.
I just do not get it how suddenly a good line is thought to achieve a 161% increase, or a bad line a 32% decrease of the old style estimates.
Is it old renluop being stupid?
-
Is it old renluop being stupid?
I think it's more like BT's estimates can't be relied upon.
e.g. I have just tried a telephone No. that is roughly 300m closer to the same cab than my home (surely a different DP?), yet the FTTC estimated speeds are identical.
FWIW, this is what my connection currently achieves (despite Plusnet tests very recently givving it a clean bill of health regarding bridge taps/crosstalk/RFI/REIN etc:-
Bearer: 0, Upstream rate = 4877 Kbps, Downstream rate = 20695 Kbps
Test Outcome Pass
Test Outcome Code GTC_FTTC_SERVICE_0001
Description GEA service test completed and no fault found but unable to check for customer equipment connected to modem.
Main Fault Location OK
Sync Status In Sync
Downstream Speed 20.7 Mbps
Upstream Speed 4.9 Mbps
Appointment Required N
Fault Target Fix Time
Fault Report Advised N
NTE Power Status PowerOn
Voice Line Test Result Pass
Bridge Tap Not Detected
Radio Frequency Ingress Not Detected
Repetitive Electrical Impulse Noise Not Detected
Cross Talk Not Detected
Profile Name 11.2M-22.4M Downstream, Interleaving Low - 2.5M-5M Upstream, Interleaving Off
Time Stamp NA
So, my connection must now be FTTC Range B (Impacted???) as I used to achieve around 30Mbps/7Mbps
-
I might be more convinced by this checker but I used the numbers of 3 neighbors. I would guess a range of distances of 220m to 350m to the CAB. Each of them nearer the cab and find the same totally identical result to my own!! ???
They probably are all from a common box in the ground, two then go underground and the other two (including mine) come via a pole standing near the box in the ground. In fact a walk along the road and judging cable runs carefully narrows the range of distances to about 280-340m so it may be luck or approximate estimates.
-
Hi
If your line has no faults to the Master Socket (dside drop wire lead in) the results should be the same on all ISPs tests.
If it is different they have place you on a different class of service. QOS even different service routing after their provision.
Please remember the distance is to the DP on Pair 1 from the Cab only and the line resistance result,this should be checked on provision of the DP block.
Regards. Phil
-
Surprising how fiercely an impacted line appears to affect performance.
There are my daughter's property and one down her road nearer ti the cabinet by 77 metres, both estimated 80 Mb/s, Clean and dirty.
Daughter's low estimates are 9.9% (clean) and 26.25% (impacted) below the full rate, whilst the other property has 2.375% and 16% reductions over full rate. That's 8 Mb/s in 77 metres for a dirty line.
-
Hi Ren
Are all the three lines with the same ISP in my thinking the results should be the same.
Regards Phil
-
Hi Ren
Please check if all three lines have the new Master Socket.(speed point)
Regards Phil
-
Whats the difference between low and high - my line syncs at around the "clean" low rate, not the high - the higher is higher than the estimates BT have ever shown for my line !
-
gusy check my new thread, basically a whole can of worms has been opened as far as I am concerned.
My new impacted estimate is EXACTLY the same as my previous estimate, that suggests my line is in a impacted state.
Greybeard who previously took interest in my line may take interest as well.
-
Sorry if I wasn't clear in that last post of mine. :-[
The details were the theoretical ones from the BTw estimates.
In real life they do in fact have fibre via PN and get the full whack, being 160 metres from the cab
-
I found that the harmonic mean of the estimates for my would-be connection were not that far from the old style estimates. I wonder what formula they may be using.
-
ok crossposting this here as I think is relevant.
Plusnet have revealed the gains of vectoring so far arent great, considering some of us have seen speed drops of around 40mbit and 15-20 isnt uncommon, plusnet reporting gains of 10 is underwhelming, and it would seem more is at play than direct crosstalk. After flashing the new firmware I now see some more data on my graphs which suggests BT are cutting power back in addition to crosstalk issues.
Pasting below.
I flashed my modem today with the new firmware. The one with btagent removed and gui added.
Made some graphs, and something interesting has come to light since these graphs now show upstream data I wasnt seeing before.
They show.
Upstream attenuation lower than adjacent downstream tones (BT artifically increasing my downstream attenuation?)
Massively lower QLN on U1 tones, easy to see why, see next.
Massively reduced SNR on U1, power cutback for sure.
SNR on U2 almost as high as end of D1. Can clearly see my U2 has more snr than adjacent downstream tones. This is even same QLN on U2 as downstream this is a reasonable indication the downstream has power cutback on it.
-
Black sheep, sorry I only just found this thread.
I actually thought you told me you never see big crosstalk before :) (apologies if you never said that)
I was one of those saying I think crosstalk is a big problem at least for some lines.
so if you want my opinion is 15meg crosstalk realistic on a line that synced at 80mbit? absolutely.
My line I have estimated roughly 40mbit lost to crosstalk down from 110 to just under 70. Plus 9mbit lost on the upstream from 36mbit to 27mbit.
Just seen this myself, after having taken a break from the t'internet.
I honestly can't remember saying I've, "never seen big crosstalk" ?? If I did, it may have been during my early days of VDSL faulting, when the learning curve was steep and the amount of VDSL-enabled EU's was relatively low ?? Ergo, low levels of X-Talk. I can't remember ??
I'm still awaiting the official feedback from the vectoring trials. There are 'Internal viewing only' (oo'er missus) 'docs' on our 'Intranet', showing various results at various stages on differing line lengths, but they are not the easiest of 'docs' to interpret. Lots of acronyms and jargon that I've never heard before ??
I'm sure something will happen sooner rather than later, as feedback to the Chief Engineer's Office gathers pace due to increased VDSL take-up ??
-
Hi. Customers
On the provision of a new instalition from the Cab D side a high and low insulation test should be taken to the Master Socket. ( speed point )
This consistes of A and B lines to Earth plus loop resistance in ohms these results should be given to the customer in their contract with there ISP.
This test will not pick up slip pairs on an unbalanced cable but these will be corrected by the special faults group in my day.
All overhearing faults if were difficult or vague were passed to SFG.
Regards. Phil
Hi Phil
I haven't heard of this myself. But with the 'Self-install' being very new to the business, maybe the protocols haven't filtered through to the ones that need them, ie: the engineers ??.
If the tests you mention are to be carried out from the Cab looking towards the Master Socket, then a very easy test for 'Split Pairs' would be the 'AC Balance' test. Literally two presses of a button and the results are there. A little more in-depth to perform but still do-able, would also be to test for 'Noise'. Both these tests would give an indication as to whether a 'Split pair' was in use ................ obviously, that is dependant on the fact that all the other MPF tests pass conclusively. :)
-
Ive posted about this else-where but I feel it also needs to be added here.
When BT first brought out the 'impacted info', my own line data still showed the same estimated speed for impacted. Both lines of data were the same. Over the past few weeks, my max headline speed has now dropped due to cross-talk, meaning that anytime now Im expecting this to have a real impact on my connection speed. Suddenly some-time over the past week, for the first time the checker is returning a different set of speeds under 'impacted'. They definitely were not there last year - Ive checked many times in the past.
Is it a co-incidence that impacted figures of 63Mb now have appeared only when the possibility to affecting my connection speed through cross-talk. My clean is still showing at 80Mb, but impacted goes down as low as 53.8 Mbps.
-
Re replies 38 & 39, can anybody enlighten me as to how such tests can be performed everywhere when there appear to be so many subcontractors who have no more than an inverter-powered modem without any access to port 2 ?
Surely there can be little point in the noble Black Sheep and his flock placing their seal of approval on the relatively small number of lines they can be involved with, leaving a nasty but undefined mess elsewhere.
I would also like to know what End User commercial conditions are applied to any achieved repairs of impacted lines.
Kind regards,
Walter
-
Nail on head, Walter. This is as much a bug-bear with genuine engineers as it is EU's and ISP's alike. There's always been, shall we say 'Issues' raised, up the management tiers as to why our headline product was being a) Installed by the junior members of staff ie: the ex-MOD guys, and b) Allowed to be repaired by severely under-trained engineers, some of whom still have only an analogue SA9083 as their means of testing.
In BTOR's defence, our then CEO (Liv) addressed the tools issue, and has made massive headway in getting engineers kitted out with the necessary. The Contractors ?? Can't comment on that as I have no dealings with them.
However, even with the right kit you still need to have a healthy understanding of the electrical principles to decipher what the instruments are telling you.
We do have some good apprentices coming through the ranks, but we have far more staff that require a better understanding of our job. I lay that particular blame firmly at the 'BT training' door.
-
@ BS,
As well as electrical principles, perhaps we could also include a detailed understanding of DLM logic**, the effects of using Huawei and ECI modems on the "opposite" MSAN equipment and the changes which are quite obviously included within each new updated modem firmware release on both platforms.
"Simples" it is NOT !!!!
** Just as a trivial example, who can tell me why some VDSL services are capped (banded ?) at integer sync speeds whilst others wander about just as the spirit takes them at any sync speed they select ?
Oh, and don't forget the joys of self install by EUs provided with standard dongle filters and who can't even recognise what a ring wire is; let alone the benefits of an integral filter faceplate.
(E.g a Sky self install with a BT Wholesale estimated 42.4 Mbps but actually achieving 13 Mbps.)
Kind regards,
Walter
-
I have to say, I don't know anybody who has a full understanding of DLM logic. Not even our resident brain-boxes have the complete picture on that one, but I take your point about 'awareness'. The only addendum I would add to that, is if the engineer has comprehensively ensured that the MPF is as good as it can be, then DLM activity is nothing to do with us ..... see: RFI, REIN, EMI etc etc .
I personally do not think self-install is a good idea, and won't be as big as some think. I probably get one a fortnight whereby the EU has 'No synch'. They have usually moved into a new premises and inadvertently plugged a micro-filter into an integrated filter. This should of course, be resolved via the 'Fault reception' front desk, but it isn't. I shan't mention which ISP tends to be the guiltiest regarding this, suffice to say they are the ones that are pushing the self-install product more than any other.
The same example would probably follow during a VDSL install with some EU's ??
Lets be reasonable here though, these ISP's aren't in business for nothing. They will (as they always do) be working on 'percentages' as a business model, in as much as most EU's will manage a self-install if they choose to, thus enjoying a cheaper 'Customer experience' :sick:. The rest will need help to maximise their installation, and will have to pay for it.
Cheers Walter.