Info for customers with fibre faults.

[#1]

Ignore the above - on further reading it looks like some kind of power cut back (PCB) mask is applied at the cabinet that cuts back certain tone groups. Who would have thought!

[#2]

Quote from: ejs

I hope I'm wrong, but I'm not really expecting that they'll be much the engineer can do.

Oh dear, nothing like being optimistic
As misterTony has mentioned, if the install was done by a "phone faults" engineer maybe something got missed or wasn't done right. One can at least hope that everything will be thoroughly checked including the Master Socket and the D-side pair and maybe there might be a better spare pair than can be tried.
That of course does not detract from the rest of your information, and should that simply turn out to be the cause, then this would be a total waste of an FTTC connection and in terms of US performance a degradation. There would be no way then this could be regarded as an "Upgrade" and if no improvement over ADSL2+ can be achieved then misterTony should be allowed to go back to ADSL2+ if desired at/with no penalty.

[#3]

AO's comment does rather make me start to wonder if many of the lower than expected FTTC speed experiences have a cause related to the distance of the cab from the exchange?  I am guilty of having believed (accepted / understood?) that the only limiting factor for FTTC speeds is the premise's distance from the cab and the concurrency of FTTC services (cross-talk) - full stop nothing more.
EJS's insight makes plain that there does need to be some other considerations; hitherto speed degradation has been put down to cross-talk from the increasing up-take of FTTC - may be now we need to consider the influence of the AVOIDANCE of cross-talk with weak-remote-from-the-exchange ADSL services as well?   It would be very interesting to assemble a body of data for analysis identifying exchange-cab distance, d-side distance and nominal download speeds.

[#4]

Quote from: Townman

I am guilty of having believed (accepted / understood?) that the only limiting factor for FTTC speeds is the premise's distance from the cab and the concurrency of FTTC services (cross-talk) - full stop nothing more.

Exactly my thoughts too before reading this. From what I read on other forums, it looks like there may be a limited number of 'PCB masks' that can be applied, but I don't know how many.
I agree that it would be very useful to get a feel for what impact these power reductions have on speed. I am not even sure how you would go about finding out. I wonder if there is any published information about how these masks are applied and under what circumstances, I will have a dig.
The impression I get from reading about this (I have only read very little so far) is that the masks reduce the power at certain frequencies (i.e. certain tone groups), rather than just reduce all frequencies equally. I assume this is because the lower VDSL frequencies may impact the ADSL more?
Either way, it is a bit of an eye opener as compared to my previous understanding.
More research required I think.

[#5]

It is an eye opener as you say, however you are correct that it should only be tones (upto 512 for ADSL2+, 255 for ADSL) that ought to be affected by such masks. But it's possible that on longer connections from the Cab to EU that those tones might otherwise carry a significant number of bits compared to the higher tones (attenuated more severely by line length as well as being possibly subjected to a wider spectrum of interference) and this results in a greater impact than might otherwise be expected on achievable speeds.
However, the fly in the ointment here in misterTony's case is the fact that the US is virtually half the speed when on ADSL2+. This suggests that something else could be amiss.

[#6]

There's also the possibility that far away from the exchange, thicker wires (giving lower attenuation) might have been used for the d-side.
I don't think much can be drawn from comparing the FTTC speeds with what the speeds were on ADSL2+. For FTTC, there's the small US0 band using the lowest VDSL2 frequencies, then there's a downstream band, then an upstream band. With most of the upstream on higher frequencies than the downstream, it's not surprising for long lines to get pretty poor upstream speeds. Considering the upstream bandwidth and that the upstream attenuation reported was far less than the downstream attenuation, perhaps the upstream is only managing to make any use of that little US0 band on the lowest VDSL2 frequencies.

Quote from: misterTony

High Low High Low FTTC Range A (Clean) 21.6 14.7 2.1       1 FTTC Range B (Impacted) 15.9 6.5         2.1       0.8

With these estimates, it should have been obvious that FTTC should have been approached with caution, with the potential for it to be no better or even worse than ADSL2+.

[#7]

Quote from: Anotherone

However, the fly in the ointment here in misterTony's case is the fact that the US is virtually half the speed when on ADSL2+. This suggests that something else could be amiss.

Thanks for all the comments on this which makes for interesting reading.
I realise that after the engineer visit on Tuesday, that there's a possibility that nothing may change. I would have to accept that and take the contentious issue of how this can be considred an 'upgrade' up with PN (something they seem to ignore and reject atm). However I do hold on to the slightest hope that something may not have been done correctly during the original install and that a fresh set of eyes may be able to eek a little more out of my connection. Time will tell either way.

Thanks again folks 🙂

[#8]

Quote from: Townman

EJS's insight makes plain that there does need to be some other considerations; hitherto speed degradation has been put down to cross-talk from the increasing up-take of FTTC - may be now we need to consider the influence of the AVOIDANCE of cross-talk with weak-remote-from-the-exchange ADSL services as well?   It would be very interesting to assemble a body of data for analysis identifying exchange-cab distance, d-side distance and nominal download speeds.

If that's what's in play here, would that mean that all end users served by that cabinet would be limited to "impacted" speeds?  It sounds to me that it should be something that's accounted for in the "clean" estimates if it's an unavoidable effect of the line lengths.

[#9]

That's an interesting point aesmith. And if the estimates provided originally don't take that into account, should that be the case, it seems to me that's just bad luck for Plusnet and Openreach. However,

Quote from: ejs

...... perhaps the upstream is only managing to make any use of that little US0 band on the lowest VDSL2 frequencies.

But, that US0 band uses the same tones as ADSL/ADSL2+ (give or take a tone) so at the moment I'm not seeing why the US would be so badly affected.

[#10]

Quote from: aesmith

If that's what's in play here, would that mean that all end users served by that cabinet would be limited to "impacted" speeds?

I am not sure. But I would hope that 'impacted' would be reserved for issues that relate to something other than power cut back being applied to the lines.

[#11]

AFAIK 'impacted' isn't supposed to include PCB.

[#12]

It will be interesting to see what comes of the OR visit.  It really seems like the U/S speed is indicative of a problem, and maybe resolving that will improve download as well.  It's possible that the references to the impacted speeds was just Plusnet trying to opt out of doing anything at all, after all their original explanation about OR modem and "line tuning" was complete nonsense.

[#13]

Quote from: aesmith

It's possible that the references to the impacted speeds was just Plusnet trying to opt out of doing anything at all, after all their original explanation about OR modem and "line tuning" was complete nonsense.

Actually what they said is kind of correct, it's just the wording could be better.

[#14]

Correct in what sense?

[#15]

Err, so why does everyone think the low upstream indicates some kind of problem (presumably that an engineer might be able to fix), rather than just a long line performing slightly below the estimates? Perhaps the FTTC upstream is lower than the ADSL2+ upstream, because there could be higher levels of interleaving, with more of the upstream being used to carry the FEC data.