btw, imgbox.com is free and doesn't have a resolution cap

In reply to a post by qasdfdsaq:

In reply to a post by RobertoS:

In reply to a post by qasdfdsaq:

I also used a shorter modem lead and wired it straight into the socket (the filtered IDCs behind the bottom plate) which also cut off about 0.3-0.5dB.

?????
Pardon?

Do you mean the un-filtered?

Nope, I meant the filtered ones.

Only the phone is filtered. The broadband is unfiltered.

In reply to a post by RobertoS:

You have me thinking I've thrown a complete wobbly!

There should be no VDSL2 signal on filtered terminals. The filter is supposed to removed anything outside audible frequencies from the output, as otherwise you wouldn't be able to use the phone because of the racket. xDSL connections are unfiltered.

I don't think I've gone mad. So either I have, or you are not correct in what you say you have done, or you need to see what happens if you connect the modem with a standard ADSL cable into the DSL socket on the front of the faceplate.

There are two sets of connectors - one for xDSL, one filtered telephone (for phone extensions), then there's the "master" connector on the NTE5 too, but that's not meant to be customer accessible.

I've (obviously) connected it to the one meant for xDSL, though whether that's xDSL filtered or completely unfiltered I'm not sure - I assumed the xDSL and phone sockets were two oppositely filtered (or two sides of the same filter) points - but if you say the xDSL is completely unfiltered, then sorry - I did mean unfiltered, I just didn't know it. I always just thought the DSL side had voice filtered out and vice versa.

Phew!

xDSL lines are unfiltered.

However, what about the phone(s) now. From memory, as I have to dog-walk, if picking up the phone causes a problem have you tried an extra normal dangly filter on the phone socket? This could be a red herring, but as I said, I haven't time to check back through the thread.

In reply to a post by Croftie:

btw, imgbox.com is free and doesn't have a resolution cap

Ah that's better (the resolution, not the stats)

72 Day Stats

I haven't tried an extra dangly filter, no. I'm not that fussed about the phone, as we only ever use it to avoid the monthly "you did not use the phone" penalty charge. Previously, my attainable rate would drop from 87 to 82, now I've improved attainable by about 5mbps, I'd expect a drop from 94 to 89, still plenty above the 80mb cap anyway.

If it does ever drop to the point it actually causes a problem, I'll take another look, for now I'm just leaving it disconnected till DLM (or an engineer) declares my line fit again.

Hehe, I've not forgot about you

It was Bald_Eagle's 4dB drop sometimes when using the phone that had come into my head.

If that had been on your connection I would have been wondering if the filter in the faceplate had given up. Hence the suggestion of adding a dangly .

In his case, it looks more like a D-side problem.

Ah, right. Yes, 4dB is quite a drop and my connection loses about 5 megabits per dB!

Bald_eagle1's line is quite odd. At that length, I'd expect about double the speed he's seeing. Some sources say even up to 80 is possible at that length, but closer to ~50 under "average" crosstalk conditions.

It doesn't help that we don't have a simple to evaluate attenuation value for VDSL2 connections (certainly not via the HG612 modem), although an engineer's JDSU does combine attenuation from the 3 DS & US bands into single DS & US values.

From what I have seen of other connections, my attenuation values right across the board would have to be quite a bit lower to achieve higher speeds.

What I unforunately do not have is any attenuation values from when FTTC was installed.

I am convinced that whatever caused my connection to go awol for 4 days last July has never been fully rectified, leaving me with higher attenuation values & therefore susceptibility to "noise interference" than I had at first.
I just can't prove it

Apparently 13.8dB per km is typical attenuation for 0.5mm twisted pair copper cables.
23dB is the lowest DS attenuation value reported from an engineer's JDSU.
Once over, 30.2dB was reported, again from a JDSU.

Edited by Bald_Eagle1 (Mon 07-May-12 22:55:40)