Nice work! I found another cab nearby on Elm Close but only the FTTC box, not the main one. Either way, these are about 0.5-0.7km away from my house, which ties in with the attenuation I am getting.

If the attenuation ties in with the distance then it certainly does point towards a "noisy line". As for the cause - there are many including cross talk which can be worse for a line underground for most of the route.

Well, it does sound like the issue is with the line from the cabinet (which is 450m away by road - less as the crow flies) is knackered/noisy etc. I have now taken everything out of the equation - the power line, by moving the wiring away from there, new VDSL faceplate, no extensions, a successful quiet test... nothing... still "slow".

Oh well, at least it's much tidier now than it was before.

I would not say "knackered" ... it is just that some lines will be a lot noisier than others due to various sources.

Larry desperate attempt before I ask for the line to be reduced to 40/10... Bought a Billion 880NL router so we'll see how it fares.

Here are my tones stats from the newly acquired Billion router:

https://dl.dropboxusercontent.com/u/43025045/stats.JPG

What does that tell me/you?

In DSLStats, select <Configuration>, then <Items to Monitor>. Under Others, tick <SNR per tone> and <Include with bitloading>

Then there will be a graph under Bitloading - which has teh SNR and Bitloading included - it tells a little more.

The dip from tone150ish is to be expected although I would have hoped it would recover by the time tone 300 is in use but that could be the power mask. I am surprised it cuts off around 1950 too.

Thanks - here are some more graphs.

https://dl.dropboxusercontent.com/u/43025045/Bitload...
https://dl.dropboxusercontent.com/u/43025045/Bitload...

Does it tell you anything interesting?

There are a few nasties in the Tone 211 to 350 region. 211 is 909kHz - a BBC radio frequency! and the other notches are too. They appear on others but yours is quite drastic. Do you live close to a radio transmitter?

What I can never understand is why the top end cuts off ... you are still seeing 6 bits per tone and even 1 bit is usable. I wonder if there is a lot of noise at the top end?

To me it looks more like the chart is cut-off rather than the bit loading. A function of how it has been scaled and captured. If the bit-loading had been cut-off, then wouldn't we just have seen those frequency bins at the top end with 0 bit loading?

nb. the low bit loading from about 210 to 350 might not be because of a local source of interference. It could easily be down to the power mask being applied, very possibly to prevent interferences to the BBC radio channel. Without knowing what power density is being used at the cabinet end, it's difficult to say. However, I'm minded to think it's primarily a function of the power masking, not interference.


*** update ***

This may be worth look at. It's the ANFP for VDSL from street cabinets (in an early form, so might not be completely up-to-date). However, what it does show as that there is a big cut in the power mask applied which is there primarily to prevent interference with exchange-based ADSL services. However, the profile varies according to distance from the exchange with the logic being that ADSL2 frequencies become more degraded with distance and, by the time attenuation drops to 68dB then the VDSL2 service might as well reuse that frequency.

See around page 27

http://www.niccstandards.org.uk/files/current/ND1602...

If the cabinet is close to the exchange, then that power roll-off can extend all the way up to about 1.5Mhz. (The result of this is, perhaps perversely you might think, the worse you ADSL2 service the better you might expect your VDSL to be for any given subloop length).

I'd also add that there may even be some "notching" applied for very specific frequencies, perhaps to minimise radio interference. I have heard talk of such things being done. That might explain the places where bit loading drops to zero.

It also means that simple broadband speed estimates based on distance from cabinet alone could be misleading, especially for slightly longer subloops with cabinets a long way from the exchange.

Edited by deleted (Wed 16-Sep-15 09:42:24)