That has to be pathetic . I can't think of a way round it though.

Agreed.

Still seems daft that exchange-based ADSL is giving 5Mbps, but it won't come up with something better for a cabinet-based setup. Perhaps it is a problem caused by power-reduction in the cabinet.

How much closer does cabinet 1 get to the house compared to the distance to the exchange?

ISTR the cab is just outside the exchange

Ah. I've lost any suitable VDSL2 graphs to compare with this one, ie straight speed against attenuation. There's at least one that I've seen in the past, I think it had it by profile. Unless that one did speed against distance.

Yeah - exchange is just about 2 miles away & cabinet is around 20 metres from the exchange.

I know obviously I wouldn't benefit that much with most of the distance still being copper but there is no ADSL2+ / LLU at my exchange so I can't benefit from speeds higher than 8Mbps... and as I'm syncing with interleaving on at 7Mbps, I should be able to expect the same speeds as LLU/WBC 21CN?

Cheers

Jack

In reply to a post by RobertoS:

Ah. I've lost any suitable VDSL2 graphs to compare with this one, ie straight speed against attenuation. There's at least one that I've seen in the past, I think it had it by profile. Unless that one did speed against distance.

I've got another of those pesky Ericsson Vectoring documents, this one a PDF of some powerpoint slides - with some of the raw research on crosstalk and vectoring.

The graph on page 9 seems most appropriate to our deployment: 0.5mm copper, with the 17a profile (the earlier graphs are for 0.4mm copper).

The red line should represent the worst case behaviour of lots of crosstalk going on - and is probably a good prediction for what the DSL checker will start reporting next week.

I guess our experiences will probably lie midway between the red and blue lines.

In reply to a post by RobertoS:

Ah. I've lost any suitable VDSL2 graphs to compare with this one, ie straight speed against attenuation. There's at least one that I've seen in the past, I think it had it by profile. Unless that one did speed against distance.

Aha - another Ericsson document, this time with the graphs you probably wanted...

Page 7 shows how speeds vary over distance for profiles 8d, 17a and 30a. Again, it seems to be the case of maximum crosstalk, but it shows that VDSL2 seems to drop to ADSL2+ performance after around 1800 metres - or just over 1 mile.

In reply to a post by jdowning640:

Yeah - exchange is just about 2 miles away & cabinet is around 20 metres from the exchange.

I know obviously I wouldn't benefit that much with most of the distance still being copper but there is no ADSL2+ / LLU at my exchange so I can't benefit from speeds higher than 8Mbps... and as I'm syncing with interleaving on at 7Mbps, I should be able to expect the same speeds as LLU/WBC 21CN?

My old line, on 20CN, was borderline 8128 without interleaving, with a 30.5dB attenuation - which Kitz predicts to be 2.2km. With 21CN it gets around 13-14Mbps sync.

At 2 miles, you will be worse, but I'd still expect maybe another couple of Mb on top of ADSL1. As I pointed out in the previous post, VDSL2 seems to run out of advantage at around 1 mile, so I wouldn't expect it to be any better than ADSL2+ for you... but that still ought to come in over the 5Mbps threshold.

Does BT use vectoring or looking to use vectoring? seem to be quite an automagical tool to help boost the speeds.

Yes. In one of their annual results documents, they recently mentioned the 17a profile change to get the "up to 80Mbps" service in 2012, and then mentioned "other technologies" to take it to > 100Mbps - but in a graphic to show future availability, they mention Vectoring by name.

They don't give a date for it though... I suspect it will have to wait until the manufacturers get a competent implementation first.