FTTC upgrade - a pleasing outcome

The engineer turned up shortly after 10 this morning for my FTTC upgrade to Zen Fibre Active.

The BT checker predicted 67.6Mbps down and 20Mbps up, though I knew it tended to be on the pessimistic side and allow some latitude for speed drops as the number of users (and hence crosstalk) increases.

Initially, the engineer couldn't locate the pair at the PCP - the records showed a pair that, in his words, was hanging in the wind. He connected his oscillator to the line and disappeared back to the PCP. I realised he'd found the correct pair when I saw my ADSL router lose sync.

He returned, carrying his JDSU, a new CTE5 (the version of the NTE5 with IDC blocks for the incoming pair), VDSL filter, HG612 modem and tools. We knew we were on to a winner when we looked at the JDSU. He let me take a photo, so I have a record:

Up is the first figure of the two.

Actual rate 20000K 80000K
Max rate 38501K 110085K
Capacity 51.9% 72.7%
Noise margin 23.2dB 14.7dB
Line atten 16.7dB 12.8dB
Signal atten 16.6dB 12.7dB


The JDSU estimates the D side length at 0.5km. I think the real figure is around 220m, but the engineer neither had the maps nor a TDR derived figure with him. He commented that it was a particularly good pair. Apparently we're the second install on that PCP - the exchange only started accepting orders about four weeks ago.


After a slightly stuttering start whilst everything settled down and I sorted out a few router related kinks (I run pfSense on a 1U rack server), all is well now - I'm getting full speed. I hope in due course to add RFC 4638 support to pfSense, so that I can take advantage of a 1500 byte MTU - this may have to wait a while.

how long ago was your cabinet activated?

as I am not sure why you think the estimate distance is wrong, line routing often doesnt take a direct distance.

Great initial sync speeds but if your cabinet has been only activated a very short time be wary.

I started off with similiar attainable and then had attainable 73/24 under a month after my install.

my direct distance to cabinet on same road about 200m as you think yours is, but my attenuations suggests nearer 500m, and my current speeds suggest more than 200m also.

Ok I see now you mentioned how new it is and you are the 2nd install, just be wary of future crosstalk and hope your pair isnt too sensitive too it (different pairs will pick up varying levels of crosstalk).

I tried to acknowledge in my first post that things will likely deteriorate over time due to increased crosstalk as more people switch to the FTTC cabinet. However, Virgin Media got here before any flavour of DSL, so a lot of the houses are on Virgin. There's never been any noticeable contention on 20CN or 21CN ADSL.


If I finish up with 65/20, that will hardly be a disaster - that is a sizeable improvement on the 17/1 I had with ADSL 2+, especially in the upstream direction. I know my speeds will likely drop a fair bit, though the JDSU figures indicate there's plenty of headroom at present. The D side pair has always performed better than expected on ADSL, but may be significantly less good at the higher VDSL2 frequencies.

I've done what I can to ensure my internal wiring is clean - there's an underground feed to the crimps, and the crimps were replaced a few years ago so shouldn't be some horrible dried out mess that's full of corrosion. From there, the line travels upstairs in a three pair cable along with another line (voice only, no DSL of any sort), through a wall and straight to the NTE5 with the modem mounted immediately above it. This three pair cable is only around 5m long and was part of a Highway install originally, so is of decent quality. The bathroom is next door, so it's some way away from any electrical equipment.

I may well look for a better quality alternative to the long untwisted RJ11-RJ11 cable that BT supply, which then engineer coiled up and left on the NTE5. This only needs to be 30cm long.

Any speed loss due to crosstalk might be reduced somewhat if BT eventually deploy vectoring.


How do my attenuation figures compare to yours? I appreciate we may not be comparing like with like, as you may be comparing figures from a hacked modem to those JDSU figures, but it would be interesting nevertheless. 500m seems too long as the cables almost always follow the pavements in this estate, which was newly built in the 1980s. The total pavement and path distance back to the PCP, which would make sense for most cables heading in this direction away from the PCP, is only around 220m allowing for every twist and turn. I'd be surprised if the actual run was more than double that length, though it's possible. The DP is only a few metres from our front door.

If you are interested I have a short (around 18 inch) RJ45 to RJ11 cable you can have, it is made of a twisted pair removed from an old patch cable.

I doubt it will make much difference, but you are welcome to it,

indeed. I agree with pretty much everything you said.

my d1 attenuation is just over 10db. I am working on the basis of around 2db per 100m. But am aware this can vary on the quality of the cable. For all I know I could be on thin ali cable and actually be at 200m.

If you know how to track the routing of the cabling then I will try to track mine.

sadly on hacked modems a overall attenuation isnt provided. So no direct comparison to the engineer tool.

Edited by Chrysalis (Wed 09-Jan-13 13:42:40)