FTTC: Not using a new line

Got myself a new FTTC connection on my relatively newly "cured" originally exchange only line. One thing I have not seen anyone ask before (as it's likely not a common scenario) is do you actually need to be using the line for all the potential improvements to the line to be made (such as lower noise margin)?

This line is not in use at present - but will be in a month or so. Does this gap pose any problems as part of the new setup phase?

For anyone interested here at the starts:

Connection: DS: 33,220 US: 8,681
Line Attenuation (down): DS1: 21.8 / DS2: 59.0 / DS3: 87.9
Line Attenuation (up): US0: 10.6 / US1: 49.9
Noise Margin (down) 6.3 / 6.4
Noise Margin (up) 6.2 . 6.2

Approx line length: 1.1km
Cabinet type: Huawei

What do you mean by "using"?

Just keeping the modem or hub connected and powered will be enough

Your stats look fine when looking at attenuation, sync and SNR.

I would query the line length, it seems a little long for what you are seeing.

Empty house so router powered on but no user data being transferred.

Could distance not look quite correct if the copper used is thicker? Being originally a long EO it could be a little thicker? 1.1km is the most direct route to cabinet.

Edited by gary333 (Wed 11-Mar-20 20:37:30)

That's fine, there will be management data flowing and it will keep looking at and tweaking performance

As MHC says, just leave the modem/router connected and powered on.

When we have been away for couple of months at a time and turned off the router, we have returned to slow speeds, not sure if it was coincidence, but we now leave router switched on and no problems.

Thanks for letting me know guys. It looks like speeds are on the up by about +0.6mb after only 24 hours with the noise margin up a fraction (which I assume might means there is more to give from this line if Now BB support 3db).

Day 2:
Connection: DS: 33,918 US: 8,784
Line Attenuation (down): DS1: 21.9 / DS2: 59.1 / DS3: 87.9
Line Attenuation (up): US0: 10.6 / US1: 49.9and
Noise Margin (down) 6.5 / 6.5
Noise Margin (up) 6.1 . 6.2

That is just natural movement cuased by differing noise levels, temperature and other environmental factors.

You will see the major changes when they happen - sometimes after just 3 or 4 days stability, and at other times it can take 2 weeks!

I am really impressed on the performance of this line - it's well over the "high" performance banding from the BT DSL checker.

Here is what BT say it should be capable of:

High Low High Low
VDSL Range A (Clean) 35 22.4 6.8 4.3 20

Yet here it is on Day 3 it's well over the high figure:

Connection: DS: 39,789 US: 8,633
Line Attenuation (down): DS1: 21.9 / DS2: 59.0 / DS3: 87.9
Line Attenuation (up): US0: 10.6 / US1: 49.9
Noise Margin (down) 6.5 / 6.4
Noise Margin (up) 6.2 . 6.1

Measuring the length of the line via driving the route and it's a minimum of 1060m away. I expected to be getting low 20's on this line so I am very happy.

I took out the "Fab Fibre" package which I am assuming is a 40/10 package thinking I'd get nowhere near maximum speed of the package

Edited by gary333 (Fri 13-Mar-20 21:08:56)

I’m now using the line as I’ve moved. Does anyone know why this line seems to be flip flopping between 6db noise margin and 3db noise margin. (80% of time 6db, 20% 3db). It’s doing this without restarting the connection at all. Connection speed appears to be at the maximum for the profile (40mb) from Now TV.

Edited by gary333 (Sun 12-Apr-20 20:40:40)

They are quite good stats for over 1KM. You may have vectoring; run xdslcmd info --vectoring on your router and paste the result.

I suspect you'd get close to 50MBit if you upgraded to a 55MBit package as this would enable the SNR to go to 3dB from those stats by the way.

Cheers. That's good to know. I have been trying to find an attenuation vs. speed calculator but I've not been able to find one. This would make it cleaer to see if it's worth paying the couple of quid extra for the higher speed package (so ta for confirming it would). BT.com still states this line (and a couple of the neighbours) should only be between 18-32mb so I should be pleased.

Using Google maps and the measuremen tool (and knowing where this line goes) it's 1045m to the cabinet (if google takes into account hills, if not it'll be slightly more. As the crow flies (diagonal) it is 797m.

I have an older house (70's vs. 90's for the one I have been referring to above) and it's only 980m from a different cabinet (ECI) and that was lucky to get 20mb. So it's certainly clear that the Huawei cabinets are better (and pressumably the thickness of the copper in my newer house is higher) as I'm doubling that cabinet.

At first I was a bit miffed that Openreach didn't route this line via the existing ECI cab as it's only 728m from my newer property. In hindsight looks like Openreach did us all on the estate a favour.

When you say run xdslcmd info, how would I do this please?

Edited by gary333 (Mon 13-Apr-20 11:42:46)

gary333

None of the estimates would take account of you having thicker than normal copper. (They wouldn't even know about it) . However it may be taken into account when you get a speed registered on the checker, Bottom line.

You just got lucky!

In reply to a post by kitcat:

gary333

None of the estimates would take account of you having thicker than normal copper. (They wouldn't even know about it) . However it may be taken into account when you get a speed registered on the checker, Bottom line.

You just got lucky!

I am only pressuming it's thicker cooper to be getting the speeds I am seeing (the wires coming out the wall in to the NTE5 aren't any thicker than usual). However, this was an exchange only line originally and I am 99.9% on the path the cable takes so if it isn't then I've really dropped on. ADSL was terrible (c.4mb)

I would have thought though that BT/Openreach would have been able to see performance from other lines on the street though. Not complaining. I expected it to be <30 due to the >1km distance from cabinet.

I'm not unhappy in any way, i'm over the moon and just got this line so I can return to Virgin and get a new joiner discount. I expected to need the 4G and FTTC runnning together.

My only question from above was if there was a reason why the noise margin flicks from 6db to 3db at random intervals, and it is doing this without dropping the connection - which from what I've seen on here seems to be rare/odd.

The noise variation is most likely due to noise source appearing at certain times and lowering the margin, but not enough to cause a resync.

Not rare at all but for most people the variation is small so noone notices.

In reply to a post by MrSaffron:

The noise variation is most likely due to noise source appearing at certain times and lowering the margin, but not enough to cause a resync.

Not rare at all but for most people the variation is small so noone notices.

Oh I see, I think I got the wrong end of the stick as to cause. So it's not something from the cabinet that is causing the noise margin to reduce from 6db to 3db, it's something firing up somehwere or other and causing the drop. Makes sense. What type of things are the usual culprit of this? It seems to be in the early evenings when I notice SNR is around 2.4db - 3db, the rest of the day it appears to be very stable at 6.1-6.3db. I am not noticing any other SNR's between 3 & 6db (so it's not dropping steadily, just in one big gulp).

Edited by gary333 (Mon 13-Apr-20 14:41:23)

Early plasma TVs used to be a typical cause. Even next door.

Something similar could be going on.

As a first step, have you bought any new piece of kit recently? Or aware of a neighbour doing so.

I don't have anything new, however I do have a cordless phone quite close by if that could be a possibility?

Most of the neighbours are quite old so being broadbrush I would guess they haven't been getting giddy at Amazon lol. All houses are detached here so I would hope it's not coming through the walls.

WiFi is oversubscribed on the 2.4ghz band, lots of competing signals with at least 3-4 AP's on each of the main 3 bands. I do have a Honeywell Evohome (which seems to be suuspetable to interference frustratingly as it cost a fortune) and Smart Meters with the Zigbee handheld monitor that is always complaining about lack of signal unless it's on the otherside of the wall from the meters.

Additional. We do have a BT visual baby monitor too.

Edited by gary333 (Mon 13-Apr-20 15:01:43)

Boring update time: After approximately a 10 minute dropped connection at 2:50am this morning on the WAN side (exactly 30 days if anyone is interested on Now BB/Sky) and not related to a router reboot as that is showing >740 hours up time on LAN I have seen a further small increase in speed but drops in upload.

My connection speed hit the 40,000kbps cap for the first time. Upload speed has only got worse from the go live being around 7,980kbps for the last couple of months and only popping in to 8,000kbps+ this morning.

Never managed to get to the bottom of the noise interference issue I was having (it's on the back burner for now) so it'll be interesting if the noise issue causes drop outs or not as this sync speed will cause noise margin to dip to less than 2dB in the evenings.

Downstream
Connection Speed (Kbps) 40000
Line Attenuation (dB) 22.2
Noise Margin (dB) 5.6

Upstream
Connection Speed (Kbps) 8290
Line Attenuation (dB) 10.8
Noise Margin (dB) 5.2

However, noticed that the line attenuation has steadily becoming poorer increasing from 21.8 to 22.2 by approx 0.1db each month. Is this likely due to the weather?

For a 1060m + line looks like I should be counting my chickens.