I have one of them junction boxes

Nothing to worry about

Speed wise for your attenuation you're on about the right speed. Getting an engineer to change the e side and d side would change the attenuation but this could be a challenge to get done.

Oddly enough we have one of those joint repair boxes too. We originally did have ADSL here which ran at about 1.5Meg.

We got a 3G modem and never had ADSL again - 3G currently runs at anything between 6Mbps and 12Mbps down and 2Mbps up. That's with 3.

We're 2.5km from the exchange with line length of 3.6km - reading your original post I wondered if you were in the same area

I have stats for a fair number of connections in the area and there is very little correlation between line length and sync rate. Certainly the highest sync rates are among the shorter lines, but by no means the shortest.

Our 3.6km line does 2Meg sync, a 3.1km one does 5.8Meg, a 2.5km one does 4Meg. It's all over the place. And could well be to do with aluminium in the circuits.

If you can put up with jitter (variable latency), AND you get a decent signal from a 3 base station, AND it isn't contended (lots of ifs there) you might find 3G does what you want.

When it's performing really abysmally:

http://speedtest.net/result/2116883665.png

When it's performing well:

http://speedtest.net/result/2130964855.png

Do "try before you buy" e.g. get it on PAYG, as it might be anywhere between useless and excellent.

How is line length measured?

Attenuation is a much better judge of actual length.

Are these speeds from speedtests, then stop doing them for the purpose of comparing lines (speedtests compare the pc,adsl,network,ISP,internet - i.e. too many variables)

You need to compare router stats and then check whether using test socket improves speeds

Line lengths measuired by physically walking the distance (you can see them, and this isn't a built up area so they don't have places to "loop around"), and information from ISP querying database.

ADSL speeds are sync speeds from the modems (have complete stats), actual throughput is even worse.

If you work backwards from the sync rates to work out the line lengths the figures cease to be believable especially when you compare near adjacent lines from the same pole on the same loop which can differ greatly.

The impression I get is that the network is so very old that once you get over about 3km line length it becomes a bit of a lottery which might be what's impacting the OP.

But you MUST take into account issues like different target noise margins, effect of peoples home wiring.

Does the attenuation of these lines on the same pole differ greatly? Or just the sync speed?

I never work out line length from sync rate, but do use attenuation. Now unknown presence of aluminium can throw a spanner in the works there, but for copper cabling it is usually consistent because of the physics.

Every day someone surfaces with poor speeds, but some wiring changes in home or tweaks to noise margin by ISP can get them better speeds.

'Network enhancement' is a legitimate clear code for an SFI. As I said, if the engineer calls they should soon see that there is improvements to be made.

Time for an update - I've pushed Sky to get an Openreach engineer out. I did a full replace of everything (again) at their insistence and they messed about with the profile again and there was little change. Subsequently I have an Openreach engineer coming out on Saturday morning and I am going to plead with him to sort the issue. Sky told me their copper line test kept failing (sometimes with high capacitance) even when there was nothing plugged into the master socket which clearly shows some sort of an issue.

I'll let you know how it goes - once again thank you for everyone's suggestions.

As has been mentioned the best way to get this sorted...

Get onto a different e side / d side. This will solve the issue better than anything else...
The E-side is the cable from exchange to your local green cabinet.
The d side is the cable from the cabinet to your house.

From the exchange to the cabinet are several different cables. All running different ways and all different lengths.
This is why your attenuation can be much higher than a neighbors. They use a shorter e side cable than you. This means that the neighbour's cable from the exchange to the green cabinet to the exchange is shorter than yours.

An engineer can change your e-side cable to the shorter one. So you then also get a shorter line, better attenuation and much quicker speeds.
In many cases there's often a better e side than even the neighbor has with better speeds. (don't get too hopeful though)

As has been mentioned ask for them to look for the best E-side / d-side...
Basically the engineer has a tool that can show them the quality and length of the e side cable...

Often a longer better quality cable can give a better speed. So leave it to the engineer.

In some cases if you live far from a green cabinet, changing the d-side can be a good thing. BUT usually the big changes are made by changing the e-side cable.

E-SIDE = CABLE FROM THE GREEN CAB TO EXCHANGE
D-SIDE = CABLE FROM GREEN CAB TO YOUR HOUSE.

AIM = get the e side cable changed over... Unless you live very far from green cab in which case d-side could be advisable.

This is what you really want (fingers crossed). I really hope it works out for you.

Sometimes there's no shorter e-sides available (which can be a massive bummer). This OFTEN happens on new housing estates when loads of new houses are built... The properties around have already taken all the good short lines up... Therefore new estates often end up on terrible e-sides that haven't been used because they're pretty dire. Hence new estates often don't get the best speeds.

Edited by ukhardy07 (Fri 07-Sep-12 01:26:38)

Lets hope he's not already on the best cable.

Had a similar issue last week, found it was poor insulation in the cabinate, remote linetest showed everything perfect, in real life there was slight contact with an unused d-side on both legs causing high capacitance.

Thanks for your detailed and helpful reply, ukhardy, My green cab is about 50 metres away so I will definitely push for the E-side cable to be changed. Does anyone know if the BT Openreach Engineer is obligated to do this if I ask for it? I suspect not but is there anything I can say that will pursuade him to do it?

We're not on a new housing estate so hopefully that is not an issue. Just out of pure interest, I assume he puts me on a different E-side cable at the green cab? Does he then have to drive to the Exchange and re-patch me in the exchange? I've often wondered how it works. I just assume it is similar to the data networks I am used to installing.

Thanks again everyone.