The max achievable is a theoretical value, extrapolated by the modem from current conditions.

It could be that another subscriber activated, added crosstalk, and you saw the effect.

It is also possible that, as the banding has reduced, the extrapolation process comes up with a different answer.

Along with WWW's comment, there could be some additional external noise from a device in your house or that of a neighbour.

In reply to a post by darrellr:

Within a week I should be back at 74-80 down/20 up.

23rd December 09:00 resync to 22400/5999
...
9th January 06:48 resync to 55000/19999
10th January No resync
11th January 06:54 resync to 60000/19999

Has there been any progress in the last couple of weeks?

I had a small bump to 66999 mbs in mid- January and then it stuck pretty much. Twice the DSLAM forced a resync but no change in headline speed and I don't have stats for the G.INP on my router.

However, two days ago I suddenly saw a massive improvement in the noise figure and maximum attainable. I went from max attainable of 80-81 to 125! This morning my banding has increased back to the original figure of 74 mbps and I'm really hopeful that it will jump one last time to the full 80.

23rd December 09:00 resync to 22400/5999
24th December 08:40 resync to 25000/7240
25th December 07:45 resync (G.INP reduction?)
26th December 09:12 resync to 27400/8496
27th December No resync
28th December 07:29 resync to 29999/9999
29th December No resync
30th December 06:58 resync to 32400/11999
31st December No resync
1st January 08:00 resync to 34999/14999
2nd January No resync
3rd January 06:46 resync to 40000/16999
4th January No resync
5th January 12:30 manual reboot - resync to 44000/19000
6th January No resync
7th January 04:56 resync to 49000/19999
8th January No resync
9th January 06:48 resync to 55000/19999
10th January No resync
11th January 06:54 resync to 60000/19999
10th January No resync
11th January No resync
12th January No resync
13th January No resync
14th January 07:15 resync to 66999/19999
....
24th March 08:45 resync to 74000/19999 (SNR 20.4 db)

Edited by deleted (Fri 24-Mar-17 11:37:43)

It's great to know that banding is now being removed by the DLM. Also, it shows that the new DLM system has been introduced, it uses the 2 day stepped approach which was shown when my line was in the downstream SNR margin trial.

Edited by deleted (Fri 24-Mar-17 11:43:53)

In reply to a post by WilliamGrimsley:

It's great to know that banding is now being removed by the DLM.

I'll just keep my fingers' crossed that this changed is a generic improvement in DLM. The "stickiness" of banding has been a particularly poor trait of DLM for the last year, and we could do without it.

Lets hope the next 6Mbps disappears.

In reply to a post by WilliamGrimsley:

Also, it shows that the new DLM system has been introduced, it uses the 2 day stepped approach which was shown when my line was in the downstream SNR margin trial.

There's no evidence of the 5-4-3dB trial/live process here, yet.

A jump in attainable speed from 80 to 125Mbps is a change in SNRM of 12dB in theory (13dB, from 7.4dB to 20.4dB, in reality) - way, way, way beyond the scope of the 5-4-3dB trial.

In the absence of any other evidence, I would say it is more likely to indicate that vectoring was activated on the cabinet.

In reply to a post by WWWombat:

I'll just keep my fingers' crossed that this changed is a generic improvement in DLM. The "stickiness" of banding has been a particularly poor trait of DLM for the last year, and we could do without it.

Haha, I bet you remember how annoyed I was!

In reply to a post by WWWombat:

There's no evidence of the 5-4-3dB trial/live process here, yet.

A jump in attainable speed from 80 to 125Mbps is a change in SNRM of 12dB in theory (13dB, from 7.4dB to 20.4dB, in reality) - way, way, way beyond the scope of the 5-4-3dB trial.

In the absence of any other evidence, I would say it is more likely to indicate that vectoring was activated on the cabinet.

I know that, I was just stating that the 2 day stepped approach is now being applied to removal of banding.

Edited by deleted (Fri 24-Mar-17 12:59:08)

In reply to a post by WWWombat:

In the absence of any other evidence, I would say it is more likely to indicate that vectoring was activated on the cabinet.

I didn't realise that vectoring was 'live'. I thought it was a coming soon thing bundled with G.Fast.

NB. Is 125 mb/s the maximum theoretical for BT FTTC lines currently?

It is live on selected cabinets, where as on g.fast it is a fundamental part of the standard

In reply to a post by WilliamGrimsley:

It's great to know that banding is now being removed by the DLM. Also, it shows that the new DLM system has been introduced, it uses the 2 day stepped approach which was shown when my line was in the downstream SNR margin trial.

I see nothing to suggest any changes to DLM. His line was banded many months ago and is still banded now. It's always been the case that banding removes itself on some lines, partially removes itself on some lines, and just plain sticks on other lines.

The "2 day step" you're referring to that the OP details in the post above yours are actions from back in January. DLM has always acted in this way.

The only changes to DLM I've seen documented are the new lower dB SNRM targets. I recall a post within the past couple weeks saying something like OpenReach are aware of sticky banding and are working on a solution. I heard that about 18 months ago and the problem has only got worse since then.