I remember back to 2017 when my local area suffered power cut off with very high flicker on/off for over 40 minutes then next day DLM has capped my line at 74Meg from 80Meg and staying there for 61 days before PN force BTw to resetted DLM via remotely and the banded has been removed. When there was a power cut, GEA test come back with a massive 20 retrains on the line that caused to be banded on the line.

In reply to a post by candlerb:

In reply to a post by ironman12345:

Noise Margin
ƒi 10.2 dB ƒh 6.4 dB

Maybe your line has a 6dB noise margin target but has been slowly going below it (e.g. due to increased crosstalk), and DLM set a 70M cap to restore it to 6dB. The power cut forced a resync at which point it applied this cap.

This, but without DLM being involved. The line isn't capped.

Likely crosstalk (new lines activated on the cabinet) have slowly eaten away at the noise margin (SNRM).
The lines max attainable (maximum potential sync speed) drops when some new lines go live on the cabinet but the line remains connected at the same speed until it reconnects.

The power cut forced that resync and resulted in the line reconnecting at a lower rate.
DLM won't be involved.

The line is synced at 71.71Mb with the default noise margin. There is no spare SNRM for more speed so DLM isn't limiting the line.

That's as good as it gets now. Crosstalk degrades most lines that aren't within a stone's throw of the cabinet due to the lack of Vectoring.

If it went off "bang", like you just pulling the power plug out, no I doubt if it would. But if it flickered on and off a few times, which you might not even notice, with all sorts of electromagnetic noise during all that, then the error counts on the line would have gone through the roof and caused a reaction from it.

A power cut that results in multiple resyncs is pretty obvious if you're home at the time.
Power going on and off quickly won't give enough time to establish sync.
A powered on modem takes over a minute to sync to VDSL2.
Add the modem boot time to that and in most cases it's over 2 minutes.

The OP indicates the power just went out like pulling a plug.
I'd notice my power coming back on for 2 minutes.

A 6db SNR does not always mean a line is not capped. Plenty of power saving technologies exist to reduce power consumption, often this can be playing with interleaving vs output power etc.

Lookup tools such as ASSIA DSL Expresse Power Management, which is leveraged by Sky as one example.

I'm aware these things are possible, but are not used.
The OpenReach FTTC DLM doesn't adjust power levels.

The only DLM line profiles adjustments are

Sync speed (banding).
SNRM target (6dB for ECI and 3/4/5/6dB for Huawei)
Error protection (Interleaving, retransmission or Fastpath)

The OP's line isn't being restricted by DLM, period.

In reply to a post by j0hn83:

A power cut that results in multiple resyncs is pretty obvious if you're home at the time.
Power going on and off quickly won't give enough time to establish sync.
A powered on modem takes over a minute to sync to VDSL2.
Add the modem boot time to that and in most cases it's over 2 minutes.

The OP indicates the power just went out like pulling a plug.
I'd notice my power coming back on for 2 minutes.

I know and have known all that for years, but I wasn't writing a training manual.

I hope you also know that a lot of strange things can happen on the mains supply for many seconds spread over several minutes before the power goes off "like pulling a plug".

I also know from decades of supporting customers in the field that broad brush statements such as that one the OP made can not be fully trusted. What time of day was it? What lights were on? What radios were on that didn't crackle.

Mains power easily seems to a householder to "go off like pulling out a plug". That doesn't mean the power had been stable for the previous few minutes.

If you care to study my original post I'd be very surprised if you can find anything there that isn't true and correct, and nothing that contradicts what you personally and I would realise when we have power cuts.

There has been no suggestion from the OP that there were multiple re-sync's, nor from myself.

I do not agree that purely based on an SNR of 6, you can conclude a line is not being restricted.

I have seen banding applied at speeds that yield SNR of 6db on most resyncs, to lines which previously had 3db noise margin targets. I did not conclude the lines was unaffected 100% by DLM.

You cannot say "period" - I broadly agree with your statement, I cannot say it's 100% the case.

I also disagree that power supply issues are always this black and white.

Edited by ukhardy07 (Mon 11-Mar-19 18:40:41)

In reply to a post by ukhardy07:

I have seen banding applied at speeds that yield SNR of 6db on most resyncs, to lines which previously had 3db noise margin targets. I did not conclude the lines was unaffected 100% by DLM.

I broadly agree, however.... 71.71Mb is not a banded figure, "period".

The line may well be banded at 74Mb but if it is line isn't being restricted by this as it is syncing below it.

OpenReach use fixed banding levels. The 1st banding level below 80Mb is 74Mb.
I believe the next banding figure below 74Mb OpenReach use is 68Mb, though I'd happily be educated on any inbetween those.

Modems sync exactly at the banding figures or a few Kb/s below.
Such as 80,000 or 79,999 or 79,997
The OP's device would show it as 80.0Mb or 79.99Mb

Given the above I say with confidence that the OP's line is not being held back by DLM banding.

I'm also very confident in saying this isn't a result of the SNRM being increased.
DLM moves the SNRM target in 1dB increments. It would take at least 2dB to get the OP to 80Mb.
If DLM has increased the SNRM target then the line has gone from 5dB to 6dB.
Each 1dB is worth around 3-4Mb on a good line that makes good use of the D3 band, less on shorter lines that only use the D1 and D2 bands.

The OP could request a GEA test from their ISP to confirm if the line is banded, and if it has retransmission (required for sub 6dB SNRM targets).
The DLM line profile shown in all GEA tests on all ISP's is 13 days old so this would need done 13 days after the power cut.
Plusnet are happy to share GEA test results and I would guess the likes of AAISP, Zen, Uno, IDNet would also be happy to share this with customers.
No such luck from many of the big players.

In reply to a post by RobertoS:

In reply to a post by j0hn83:

A power cut that results in multiple resyncs is pretty obvious if you're home at the time.
Power going on and off quickly won't give enough time to establish sync.
A powered on modem takes over a minute to sync to VDSL2.
Add the modem boot time to that and in most cases it's over 2 minutes.

The OP indicates the power just went out like pulling a plug.
I'd notice my power coming back on for 2 minutes.

I know and have known all that for years, but I wasn't writing a training manual.

I hope you also know that a lot of strange things can happen on the mains supply for many seconds spread over several minutes before the power goes off "like pulling a plug".

I also know from decades of supporting customers in the field that broad brush statements such as that one the OP made can not be fully trusted. What time of day was it? What lights were on? What radios were on that didn't crackle.

Mains power easily seems to a householder to "go off like pulling out a plug". That doesn't mean the power had been stable for the previous few minutes.

If you care to study my original post I'd be very surprised if you can find anything there that isn't true and correct, and nothing that contradicts what you personally and I would realise when we have power cuts.

There has been no suggestion from the OP that there were multiple re-sync's, nor from myself.

I do not know the reason , all I know is that the power to the house went off.
If I had pulled the modem/router power supply plug out , it would have been the same.
After about 5 minutes the power returned and that is when the lose happened.
So It could have been re-synced at that time.
Anyway I can use the line and it might recover in time.
Thanks for all the posts.

I broadly agree, however.... 71.71Mb is not a banded figure, "period".

Good point, this had not crossed my mind.

I'm also very confident in saying this isn't a result of the SNRM being increased.
DLM moves the SNRM target in 1dB increments. It would take at least 2dB to get the OP to 80Mb.
If DLM has increased the SNRM target then the line has gone from 5dB to 6dB.

Good point also, I stand convinced. Nothing you have said is "news" but honestly I had never really applied it in practice, so some good takeaways for future.

I'm afraid my post, which was a reply to j0hn83, can seem to read as if I didn't trust what you said. That wasn't my intention, so I apologise.

It was more of a general comment that although all of us say honestly what we experience when things happen to us or when we witness a scene, we may not be entirely accurate about the actual facts of the event. Facts that we very possibly wouldn't be aware of.