Print

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

[gt94sss2]

Hi,

Sorry for so much detail in this post, but I would be grateful if those in the know could take a view at my line stats on MDWS (same username) and let me have any thoughts about my line.

It is less than 500 metres (470ish)from the cabinet, all copper and the DSLChecker reports:

                           High   Low   High   Low            
VDSL Range A (Clean)   77.5   55   20   18.2   
VDSL Range B (Impacted)   67.2   41.6   20   12.6

Since my line has been active for several years now, one would have thought that the estimates were relatively accurate - it was also one of the first ADSL lines installed on the exchange which was back when BT insisted lines had to be less than 41dB and <160nf from the exchange if that makes a difference.

The cabinet is G.Fast enabled/active since earlier this year but we are too far to receive it hence wanting to get the most of this connection.

Some history:

1. We had BT Infinity on the line running at 40/10 and then 55/10 with SNR to spare for several years. Attainable speed was usually around 65/20.
2. Last March an alarm installer connected some extension wiring to the socket incorrectly when I wasn't here. Speeds dropped below 55/10.
3. The extension wiring was disconnected around October 17 and went back to 55/10.
4. Upgraded the line to the 80/20 service, earlier this month and while upload speeds improved, the download sync rate fell even after g.inp was re-enabled on the line
5. BT Retail called this an early life failure and booked a 'boost' engineer to come out today.
6. Nice Polish gentleman turned up - replaced the Mk3 with a new NTE5c and tried testing the wiring in some of the chambers - the sync/attainable on my line has improved and is now higher than 55MB but only to 58/59MB.

Now, I appreciate that crosstalk etc. leads to lower rates in the long run but is there anything else I can do?

The 80/20 upgrade was arranged by the BT staff on their forums as it should have happened last autumn automatically but didn't and they are the ones who classed it as an early life failure and booked the engineer to come out.

Notes for engineer
Boost Template
 Customer Reports: slow connection
 Problem Start Date (DD/MM/YY): 12/02/2018
 Fault Type: Slow speeds - New fault
 Current Sync (Mbps): 53
 Current throughput(Mbps): 47
 POS (Mbps): 55
 OD Checks: yes
 Connected Devices: yes
 Is this a recent provide?: Yes
 Is this a complaint?: Yes
 Appointment History: No
 Hazards: This customers BT Infinity profile was upgraded to 80/20 however the download speed has dropped and the upload has increased from 10 to 17. the potenital on the for this line on the 80/20 profile is 78 so a dlm reset may be required

At the start, both I and the engineer agreed that this caused by DLM issue, though after he left I realised that he hadn't reset DLM at the end of his visit (after all the resyncs)... well that and the attainable speed had reached 63MB at one point on MDWS after one of his tests.

Thoughts gratefully received.

[skyeci]

what is your actual sync figure and attainable at the moment?

[burakkucat]

I would not like to hazard a guess as to the problem . . . without a full "circuit recalc" having taken place, following the many service interruptions that have now occurred during the testing.

[gt94sss2]

what is your actual sync figure and attainable at the moment?

After the work today, its now

SYNC:   58306   ATTAIN:   59350   

Both I and the engineer were expecting it to be much higher..

[gt94sss2]

I would not like to hazard a guess as to the problem . . . without a full "circuit recalc" having taken place, following the many service interruptions that have now occurred during the testing.

burakkucat, thanks - by  "circuit recalc" do you mean just leaving the line alone to stabilise or switching the modem off overnight or something else?

[burakkucat]

. . . by  "circuit recalc" do you mean just leaving the line alone to stabilise or switching the modem off overnight or something else?

Sorry, I was being brief and used "Openreach speak". Translated that means resetting the circuit to a "just deployed" state, where the profile is "wide open" and the DLM process waits for a full 24 hours before it takes any notice of how the circuit is behaving.

[RealAleMadrid]

Otherwise known as a DLM reset.

[gt94sss2]

Ah, the one thing the engineer didn't do today - will have to see if BT Retail are willing to request one when they call me tomorrow. I don't their frontline staff have access to request remote DLM requests yet.

[gt94sss2]

Apologies for taking a while to update this thread.

When I spoke to BTR, they could see the disconnections caused by the engineer but didn't think that this would result in capping so no need for a DLM reset (and they would have been unable to do it without an engineer visit)

As such, I have left the line running and note that this morning, DLM intervened to move SNRM to 5.4 from 6.4. INT has reduced from 8 to 4 but INP has increased to 57 from 48.  Sync speed has increased but not by as much as I thought it might.

Grateful for any thoughts on my line and for any guidance over how long DLM tends to take to reduce the SNRM on a stable line from 6 to 3 once its started intervening

[ejs]

Sounds like it has been put on a "Retransmission High" profile, so despite the increased sync speed, you'll have a lower IP Profile percentage and may not really gain very much if anything overall.

[NewtronStar]

how long DLM tends to take to reduce the SNRM on a stable line from 6 to 3 once its started intervening

It normally takes 2 days for the SNRM to decrease by 1dB on XdB

[tiffy]

Apologies for taking a while to update this thread.

When I spoke to BTR, they could see the disconnections caused by the engineer but didn't think that this would result in capping so no need for a DLM reset (and they would have been unable to do it without an engineer visit)

As such, I have left the line running and note that this morning, DLM intervened to move SNRM to 5.4 from 6.4. INT has reduced from 8 to 4 but INP has increased to 57 from 48.  Sync speed has increased but not by as much as I thought it might.

Grateful for any thoughts on my line and for any guidance over how long DLM tends to take to reduce the SNRM on a stable line from 6 to 3 once its started intervening

From historical observations of my lines performance I would agree with "ejs" that you are now on DS retx high profile, the most definitive guide to retx low/high state appears to be "B0 INP Rein" parameter, will likely have changed from "0" to "1" at the re-synch..
Your DS data throughput ratio will now be approx 91% as apposed to approx 97% on retx low profile.

As "ejs" has said, increased DS synch speed but very little gain if any on DS data speed.

[gt94sss2]

Thanks for the replies - yes, the line does now seem to be on a 'DS retx high' profile with 'B0 INP Rein' now at 1.

Will this change in future - for instance, as DLM continues to lower the SNRM?

If not, why reasons are there for a line to switch to a "Retransmission High" profile? My line seemed pretty stable on the downstream before this change..

While I remember:

When the Openreach engineer checked the line, he said it was about 470 meters from the cabinet and had been hoping to get higher speeds than he achieved during the boost visit. According to him a chamber at the top of our road (220 meters from the PCP) was only getting 73MB which also suprised him.

Both of these speeds seem low, especially as the engineer said the lines were all copper - its difficult for me to believe the Wholesale checker is so far out in its estimates re: top attainable speeds*

(*I have once seen 80MB attainable at this address some time ago but think thats as the HG612B I was then using reconnected very quickly following the DSLAM being rebooted)

[gt94sss2]

It normally takes 2 days for the SNRM to decrease by 1dB on XdB

Its been longer than that now on my line but it hasn't yet fallen to 4dB. I will leave it for a few days and see if anything changes.

Do we know what DLM looks for that might stop it lowering XdB further?

[burakkucat]

Do we know what DLM looks for that might stop it lowering XdB further?

I have always assumed that if following a transition of the target SNRM from N dB to (N-1) dB the circuit is then observed to be unstable, the last transition is reverted and the incremental lowering process is halted.