It would be far cheaper to buy a few satellite broadband connections and share those instead...

the power masking requires that the injected power is the same as that of exchange based services arriving at the cabinet, unless the frequency band is unuseable from the exchange in which case the FTTC can "re-use" the frequency at full power.

So a cabinet a long way from the exchange should have a good range of frequencies at decent power levels, good for long lines, but one close to the exchange is as you say hampered.

I do wonder if these things get setup correctly though.

well I am just guessing based on bitloading graphs.

Exchange based services have a lesser impact and achieve higher bitloading on the affected tones, and consider also that those tones are weaker (tones above 250 weaker than below) then it looks to me that there is a higher penalty when the cabinet is further away.

This may well be deliberate because long adsl lines may need more help from vdsl power cutback to stay working than short adsl lines.

In reply to a post by MaddogBattie:

Any other magic ideas on what we can do? On demand FTTP is going to be too expensive individually being that it is 2km to the cabinet but there is the potential of fibre being brought into the village then shared. Is this something likely to be offered?

It would be interesting to see how much BT would charge for installing FTTP.

You could ask Gigaclear just what they would need for a viable project. Where I live Gigaclear want 190 commitments (30%) over quite a spread out area.

I think it's a curve, if the cabinet is a very long way from the exchange there isn't much of a constraint on FTTC injection at all. As you get closer the FTTC has to back off to the same level of signal at the PCP as the exchange achieves (approx) and hence is more compromised.

So if you have a long line to the cabinet it's best if the cabinet is also a long way from the exchange. In fact in general there's more tones & power available on distant cabs, the benefit is not just to the long lines from the cab.

In reply to a post by ferretuk:

It would be far cheaper to buy a few satellite broadband connections and share those instead...

They'd all need to be unlimited. The ongoing costs would soon catch up with the heavy initial cost of an FTTPoD solution.

My cab is a long way from the exchange and I'm quite a long way from the cab. Power back-out blows a very obvious whole in the lower frequencies and heavily effects my transmit speed. The bit loading reflects the PBO mask, which isn't a straight forward curve.

The power back-out masks are published somewhere, I can find them if need be, but they definitely effect users in the "long long" category as the power is backed out to avoid interference with ADSL on the long d-sides.

IIRC there are 5 different masks that are applied to long d-sides, and which mask is applied is dependant on the electrical length of the cabling from the cab to the exchange, the further it is the more the attenuation applied.

Or something like that

Edited by deleted (Thu 07-Feb-13 08:43:02)

indeed, I have the graphs. The five illustrated masks are just examples from the provided calculation procedure.

With a 50 dB loss at 300 kHz from exchange to cabinet there is no penalty at 550 kHz and above compared to any other cabinet location. With only a 20 dB loss from exchange to cabinet there is a penalty up to over 2 MHz, so the 50 dB cab can put out more power between 550 kHz and 2MHz but less below 550 hence my speculation that it may be better to be on a distant cabinet.

yes but the shorter E sides have no penalty below 250 whilst the long E sides do.

The penalties applied to short E sides seem less severe (higher bit loading), this combined with the fact they get full power on the valuable 0-250 range.

Edited by Chrysalis (Thu 07-Feb-13 09:38:29)