That is horrific ...

In reply to a post by WWWombat:

I'm away, so can't look properly, but the key here looks to be in the Quiet Line (QLN) graph.

That shows a lot of noise being picked up in D3 - more than other parts of the spectrum. Normally, D3 is about as noisy as D2 or less, not more.

That explains why SNR per tone is lower in D3 than normal, and therefore why bits per tone is lower there.

Could that mean Zarjaz is right on the split pair? Would that fail to reject noise in D3 more than lower frequencies?

In case MyDSLWebStats doesn't show the username for whatever reason, this is the QLN graph for anyone interested: https://i.imgur.com/jQZTCOF.png

There definitely seems to be a problem with noise in D3, although I've no idea what would cause that.

I dialled 17070 and did a quiet line test while I was there the other day and noticed a very faint hissing - that may just be coincidence but would that be an indicator of noise at the higher tones?

Could it be caused by internal noise or is it likely to be a problem with the line itself? When the engineer came round, he indicated that the issue was caused by my DSL cable between the NTE and the modem. I've since replaced the NTE and the modem, but it appears that isn't the problem.

In reply to a post by Toonshorty:

When the engineer came round, he indicated that the issue was caused by my DSL cable between the NTE and the modem. I've since replaced the NTE and the modem, but it appears that isn't the problem.

What about the cable between them?

It sounds like a load of bull from the engineer anyway, but if it is the cable between A and B, replacing A and B won't fix it.

In reply to a post by RobertoS:

In reply to a post by Toonshorty:

When the engineer came round, he indicated that the issue was caused by my DSL cable between the NTE and the modem. I've since replaced the NTE and the modem, but it appears that isn't the problem.

What about the cable between them?

It sounds like a load of bull from the engineer anyway, but if it is the cable between A and B, replacing A and B won't fix it.

Sorry, that was me being an idiot. I haven't replaced the NTE - I replaced the DSL cable between the NTE and the modem with some CAT5e.

I've got another engineer booked in for Friday and Uno have asked Openreach for a full reason as to why the line is reverting to 40/10 and not for them just to perform another DLM reset and have it return to 40/10 overnight.

I'll show the engineer the QLN and tone graphs, although they may ask me to report that as a separate issue, depends on the engineer I guess. The two may be related, but I'm not sure.

Edited by Toonshorty (Tue 18-Apr-17 15:05:00)

I'd be surprised if the average engineer knew what a QLN plot was, or anything related to bits per tone.

or anything related to bits per tone.

Despite there being a bits per tone graph shown on every single manual test carried out on their JDSU/Exfo ?

Whether they bother to look at said graph, that's a different matter.

So the engineers turned up today and did some more investigation, they claimed our line build (whatever that is) was set to 40/10 and that they will have a new build that fixes it at 80/20 before the line resets to 40/10 again.

I've not heard of line builds before, but what I can see is that the HG612 is now showing the following stats:

xdslcmd info --stats
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 21421 Kbps, Downstream rate = 73608 Kbps
Bearer: 0, Upstream rate = 19999 Kbps, Downstream rate = 73839 Kbps

Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.1 7.4
Attn(dB): 16.4 0.0
Pwr(dBm): 13.6 7.0

Strangely our sync rate is actually marginally higher than the attainable which is odd, but G.INP isn't currently active so I suspect that will come into play soon.

I wasn't around but my brother did mention the noise issue to them, they tested for radio noise and apparently the Sonos bridge was causing tons of noise - so looks like I'll be moving that away from the modem. That may have been the cause of the QLN bump in D3 but we'll see.

So the line resynced this morning which has been a bit of a mixed bag.

The good news is that G.INP was activated and the attainable rates are now 77068/19718.

The bad news is that the actual sync rate is 40000/9999. Interestingly the downstream sync is 40000 rather than the usual 39999 it would get stuck at - not sure if that means anything.

Whatever the engineers did obviously didn't work.

i have this problem 11 months now with talk talk openreach do not know what they are doing