FTTC DLM query

[Andrew600]

There haven’t been any drop outs except for the restart for a firmware update in the early hours of yesterday, but yes noise margin has gone back up to 6db. That said the couple of months I had with 80mbps down, but only 74mbps maximum data rate didn’t make much sense, so I guess I was just lucky to get that.

The theory, or hope, is that because the fault in the copper line is halfway between the exchange and cabinet, disconnecting it will remove the voltage from cabinet together with the noise it’s generating. The line from the cabinet to master socket tests as perfect apparently.

The broadband should, I hope, then return to full speed. The last Openreach engineer I spoke to thought it was the solution given they are never going to replace 3 miles of copper to a very rural location with the switch off in sight anyway.
I’ve also spoken to someone who was affected by the same fault, served by the same cabinet, but much further way from it than I am who switched to EE with DV and they have gone from roughly 55mbps to 76mbps as a result, so I’m hopeful. Fingers crossed.

[imjolly]

if you just stay connected without any drops/resets then DLM will drop noise margin back to 3db giving you the speed increase back as before.  could take few weeks

[licquorice]

@imjolly see my edited post 14 in the thread.

[Andrew600]

An update.

I got the landline changed to digital voice on 1st May and I think it’s solving the problem. I say think because I don’t know if this would have happened anyway.

Initially there was no change. DLM wasn’t giving this one up easily.

Time line

1st May, sync speed 71mbps, SNR 6db

2nd June, sync speed 74mbps, SNR 5db

16 June, sync speed 77mbps, SNR 4db

Its definitely looking promising. My actual speed to device now via Ethernet is lower than I expected at 68mbps according to every speed test including BTs, but I will wait until SNR drops to 3db before worrying about that.

I do have a suspicion though.

Because the master socket is halfway up the stairs in a WiFi black spot with no power sockets anywhere near it, I have a long RJ11 cable running from the master socket to the hub, 20 metres I think, which has been there for at least 5 years, maybe more, so it could be damaged or degraded and will be my first target if necessary, but that’s not happening yet.

Thanks for listening to my thoughts.

[pippincp]

At least you can now be assured that the3 miles of cable from the exchange caonnot cause you a problem.

Your digital phone line and FTTC connection is now supplied exclusive via FTTP  from the cxabinet. Now you only have the copper line from the cabinet to worry about;.

[Andrew600 · Accepted solution]

It finally got there.
9 months after the initial event caused by an Openreach error and repair gone wrong, and 7 weeks after the switch to Digital voice.

Sync rate is now 79.999Mbps and noise margin at 3db.

Once DLM decides to lower the noise margin to 5db, which it seems reluctant to do, the remaining steps seem to  happen relatively quickly.
I do wish it reacted quicker to line improvements. It’s frustrating that you have no idea if line improvements, i.e switching off the copper landline, has made any difference for weeks upon weeks.

[licquorice]

Got there in the end!! Shame it took so long.

[Andrew600]

Yes, it has taken too long to be honest, but at least I never suffered in terms of actual usability. 

Thanks for your help along the way @licquorice. It was very much appreciated.

[Andrew600]

One more fairly minor detail that may help others in this situation.

The speed to hub has been consistently lower than expected ever since DLM first reduced the noise margin. For example, when it got to a 3db margin, I got a sync speed of 79.999Mbps, but a maximum actual speed of 71Mbps. Thats 3Mbps lower than before the line faults. That discrepancy has been apparent at every noise margin step.

The hub re synced again overnight and I now have an actual speed of 74Mbps, which is back to the best I’ve ever had.

I have no way of checking, but I’m guessing the ip profile hadn’t caught up and now has, but have no idea why it took an extra 2 or 3 days.