Post your FTTC line stats & unlocking info

Soldato
OP
Joined
14 Oct 2003
Posts
13,451
Location
South Derbyshire
Hi Cjd, interesting project. Have you thought about implementing vectoring?

Current Dslams don't support vectoring out of the box (unsure of this Dslam chipset) but the controller cards will need an update - hardware or at the very least a firmware update.

what brand of modem have you ordered?
 
Associate
Joined
3 May 2013
Posts
13
Hi,

Yes absolutely. Vectoring is at the forefront of our watchful eye. As Locky correctly states wewould need a card u/g. Although the g.vector hardware has been pretty much designed and is ready, the standard I believe that for widespread use in the UK, its still undergoing testing.

http://www.ispreview.co.uk/index.php/2013/04/bt-confirm-first-uk-trial-areas-for-superfast-broadband-fttc-vectoring-tech.html

Its annoying because we had to order a card without vectoring as its not quite ready! We will definitely replace as and when zyxel make one available....
It seems to be a mystery to everyone on how much extra speed it will give us but looking around the globe at test results it seems pretty promising....
Virtually eliminating FEXT and NEXT and having good quality .5mm copper lines (ours are all CW 1128 unarmored and ducted on the D side (about 8y old) I am hoping our offered speed range would be a more respectable 100 to 50 meg over our 500m to 950m range we have.

There is a number of 50 pair cables leaving our PCP on the D side so I suppose its how many other VDSL disturbers there is in your pair. Our trials so far have found our ADSL on park has no detrimental effect on the VDSL but the VDSL does affect the ADSL. To comply with OFCOM we will have to apply a given DPBO (downstream power back off) profile in the 0-1.1Mhz range to minimize this. We don't need to worry between 1.1-2.2Mhz as we have no ADSL2+ on site!

Also FYI we are all between 3.3km and 3.8km from the exchange cable length and people currently get a range of 3.5-6.5Meg so even without vectoring yet we are hoping for an uplift (written as a multiplier in BT's FTTC PCP Spreadsheet as a multiplier) of 9 or 10.

Locky, the modems we will have and will be distributing are Zyxel VMG1312 routers which yes - all report stats without any unlocking!! We will be putting a preconfig on them so its an easy plug and play self install... we are currently still researching the following two issues....

1. The best filtering to use... obviously there is the iplate but if all internal wiring is isolated pretty much any ADSL filter works well.

2. The fact we may have to replace incorrectly installed CCA or CCS cable for the last few meters in peoples properties.
Unconfirmed at the moment but...
If people will sign to a 24 month contract there will be no initial charge. Alternatively they will have to buy their modem which we believe is reasonable however we may make a small charge for sorting out / optimising the path to their master socket if they wish or it is needed...
 
Soldato
OP
Joined
14 Oct 2003
Posts
13,451
Location
South Derbyshire
Vectoring should restore the lost speed/attainable rate lost to crosstalk.

people with short/quality D-sides should see the max VDSL2 17a can provide (if BT offer higher speeds)
 
Associate
Joined
3 May 2013
Posts
13
MilanoChris,

It very much depends on the physical characteristics of your line.
Vectoring (or the g.vector standard) as you probably are aware implements complex mathematics, taking into account noise and crosstalk characteristics of EVERY VDSL line on the DSLAM. Just like noise cancelling headphones, it then applies inverse versions of that noise in order to effectively cancel that out...

Now...

If your line happens to leave the PCP as a single pair cable - for arguments sake, say it goes overhead on poles or for that matter even underground, on a unique route to your house - not alongside anybody elses line, then vectoring will have virtually no effect - possibly even none.

Now in our situation, our PCP is at the end of 250m of lane on the entrance to our estate. EVERY one of the 270 lines travels for 250m before starting to split into other ducts. Now these 270 lines are not in one cable but in a series of cables - therefore if your line travels like ours - in a cable alongside other users (which is more likely), then you should see some uplift...

It is near impossible to count the exact effect of vectoring - it seems to be a lets give it a go and see - type thing at the moment. No doubt, in 12 months time, more accurate figures will be available...

Various graphs available on the internet show that you should get your full 80Meg Download and about 15 Meg upload at up to 600m loop length...

It is worth also mentioning that vectoring mainly improves download sync and doesn't seem to have a significant affect on upload - in particular over longer distances...

In a nutshell - for most people, there will be a significant or at least noticeable uplift...

Do you know physically where / how your line runs, the cable type and its length (sorry if you have already posted this at an earlier time)...?
 
Man of Honour
Joined
20 Sep 2006
Posts
34,361
The cab is 0.6 miles away from my house along the road. I'd say direct distance is 0.5 miles, so a rough guess would be around 700 meters. My stats are as follows:

Code:
# xdslcmd info --pbParams
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason:	0
Max:	Upstream rate = 9182 Kbps, Downstream rate = 52384 Kbps
Path:	0, Upstream rate = 9264 Kbps, Downstream rate = 45222 Kbps

Discovery Phase (Initial) Band Plan
US: (0,95) (880,1195) (1984,2771) 
DS: (32,859) (1216,1959) (2792,4083) 
Medley Phase (Final) Band Plan
US: (0,95) (880,1195) (1984,2771) 
DS: (32,859) (1216,1959) 
       VDSL Port Details       Upstream        Downstream
Attainable Net Data Rate:       9182 kbps         52384 kbps
Actual Aggregate Tx Power:        5.2 dBm          12.7 dBm
============================================================================
  VDSL Band Status        U0      U1      U2      U3      D1      D2      D3
  Line Attenuation(dB):	 2.6	 37.3	 56.2	  N/A	 17.7	 47.5	 73.9	
Signal Attenuation(dB):	 2.7	 37.2	 54.5	  N/A	 17.7	 47.5	  N/A	
        SNR Margin(dB):	 6.1	 6.0	 6.0	  N/A	 6.0	 5.9	  N/A	
         TX Power(dBm):	-3.6	-4.7	 4.0	  N/A	 11.1	 7.4	  N/A	
#

Code:
# xdslcmd info --show
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason:	0
Max:	Upstream rate = 9172 Kbps, Downstream rate = 52512 Kbps
Path:	0, Upstream rate = 9264 Kbps, Downstream rate = 45222 Kbps

Link Power State:	L0
Mode:			VDSL2 Annex B
VDSL2 Profile:		Profile 17a
TPS-TC:			PTM Mode
Trellis:		U:ON /D:ON
Line Status:		No Defect
Training Status:	Showtime
		Down		Up
SNR (dB):	 6.0		 6.0
Attn(dB):	 0.0		 0.0
Pwr(dBm):	 12.7		 5.2
			VDSL2 framing
			Path 0
B:		51		238
M:		1		1
T:		64		31
R:		12		16
S:		0.0366		0.8206
L:		13994		2486
D:		885		1
I:		64		255
N:		64		255
			Counters
			Path 0
OHF:		120092900		1090801
OHFErr:		3507		6
RS:		678903437		1584945
RSCorr:		420929674		56
RSUnCorr:	134181		0

			Path 0
HEC:		33259		0
OCD:		1300		0
LCD:		1300		0
Total Cells:	3070721781		0
Data Cells:	478653035		0
Drop Cells:	0
Bit Errors:	0		0

ES:		1005		6
SES:		5		0
UAS:		16		16
AS:		282308

			Path 0
INP:		3.00		0.00
PER:		2.34		12.71
delay:		8.00		0.00
OR:		81.99		20.12

Bitswap:	76398		148

#

The cab serves a lot of houses, I'd imagine the cable is bundled together with others and the breaks out on each close. It's an underground cable.
 
Soldato
OP
Joined
14 Oct 2003
Posts
13,451
Location
South Derbyshire
Looking at your stats your punching very well given your attenuation

Thing is without seeing stats from when your cab was empty, we have no idea what vectoring will bring.
 
Associate
Joined
3 May 2013
Posts
13
Locky has hit the nail on the head.
Basically vectoring is not a speed booster - its a speed maintainer - so it ensures that your speed is not affected (or hardly affected) by other 'disturbers' to us the correct terminology... So in theory with vectoring, your line will act as if it is the only one on the cabinet!

Our testing figures here for our VDSL project show crosstalk as being the main issue. If we enable Fastpath on our DSLAM with two lines running along the same cable, they do not stay synced, so we are forced to use interleaving lowering the speed. Even over 150m, a few meg is knocked off right away...

I agree, your stats are good - must have a decent line with low noise / probably CW1128 .5mm copper with minimal joints if the DLM has now stabilised at this...

How long have you had FTTC and have you noticed a gradual decrease in sync??
 
Man of Honour
Joined
20 Sep 2006
Posts
34,361
It was installed 2 weeks ago. Plusnet put me on 40/10 as the predicted speed was 26 down. I checked the stats and asked them to be raised, I was seeing around 45-50. But on my TBB monitor I was seeing minor packet loss. I had to do some rewiring of power at home so the modem was restarted and when it came back up interleaving was on and the packet loss was gone. It seems stable now at around 43/8.
 
Associate
Joined
18 Sep 2012
Posts
130
Location
Not in the UK
Associate
Joined
3 May 2013
Posts
13
Lhw455,

The modems we are using seem pretty good too..

We have had a bulk deal direct with zyxel including other equipment but I noticed they are for sale here...

http://www.broadbandbuyer.co.uk/Shop/Specifications.asp?ProductID=15196

Can view stats on these too. Don't know whether BT do this but we are enabling ADSL2+ fallback support so if the VDSL line goes a bit wibbly wobbly ice below 20meg then the system notifies us and temporarily switches the user to a 24meg ADSL2+ service using just the 2.2Mhz band with 1Meg UL... These modems support this too... We know that everybody can get this on our village estate so this is our baseline worstcase speed for the furthest houses/customers while we wait for vectoring! Of course we are hoping for a 30/10 minimum to start with the closest possibly getting 60/15....

As I say, as soon as lines go live I will post and build up a big table of all lines showing full stats plus loop length and physical route...
 
Associate
Joined
26 Jun 2012
Posts
294
Location
Caldicot
I know this thread is mostly for FTTC, but i was hoping you could have a look at my ADSL 2+ connection and tell me if there is anything out of the ordinary
Code:
Uptime:	0 days, 2:04:49

DSL Type:	ITU-T G.992.5

Bandwidth (Up/Down) [kbps/kbps]:	1,276 / 12,687

Data Transferred (Sent/Received) [GB/MB]:	1.99 / 829.91

Output Power (Up/Down) [dBm]:	12.6 / 0.0

Line Attenuation (Up/Down) [dB]:	11.2 / 19.5

SN Margin (Up/Down) [dB]:	5.9 / 5.8

System Vendor ID (Local/Remote):	TMMB / ----

Chipset Vendor ID (Local/Remote):	BDCM / IFTN

Loss of Framing (Local/Remote):	0 / 0

Loss of Signal (Local/Remote):	0 / 0

Loss of Power (Local/Remote):	0 / 0

Loss of Link (Remote):	-

Error Seconds (Local/Remote):	12,178 / 5

FEC Errors (Up/Down):	0 / 5,382,768

CRC Errors (Up/Down):	0 / 110

HEC Errors (Up/Down):	0 / 907

Usually get around 10.5-11Mb down and 1 Up
 
Associate
Joined
3 May 2013
Posts
13
ConnorJoshua..

I obviously don't know much about the physical characteristics of your line but looking at the attenuation your actual ADSL line back to the exchange is about 1400m...
Your SNR is on its target at around 6dB (just a little below which is fine) suggesting the modem is squeezing what it can out of the line reliably with just that few db of overhead for any impulse noise... however....
Your upstream is maximum for ADSL but your downstream seems a little low for a 19dB line... according to real world data I would be expecting around 19 meg on this length....
One possibility is there could be a higher level than normal of background noise on your line... by the looks of it - possibly a constant noise? ADSL2 uses frequencies up to 2.2Mhz so anything up to that??

Do you know the path of your line back to the exchange?
Have you tried the normal... isolating all internal telephone wiring, plugging into the test socket on the NTE5, and trying different filters?
Does your internal or external cabling run near any sources of electrical noise?
Does your line stay synced reliably or do you have regular re-syncs?
 
Soldato
OP
Joined
14 Oct 2003
Posts
13,451
Location
South Derbyshire
If you wouldn't mind looking what chipset they are, his ADSL has a crap load of FEC erros and error seconds for a 2hr uptime, there is an issue with his line at some point.
 
Associate
Joined
26 Jun 2012
Posts
294
Location
Caldicot
Do you know the path of your line back to the exchange?
Have you tried the normal... isolating all internal telephone wiring, plugging into the test socket on the NTE5, and trying different filters?
Does your internal or external cabling run near any sources of electrical noise?
Does your line stay synced reliably or do you have regular re-syncs?


I believe the line back to my exchange follows the main roads, which comes to around 0.9 miles or 1.44Km.

Tried the test port on the master socket and that only resulted in a better ping. We do connect the router to the master socket via an extension which is plugged into another extension (God knows who wired this).

As for electrical noise our line does enter a few foot from the fuse box, i don't know if that would make a difference. There is also a large transformer looking thing down by the cabinet.

As for resyncs, you can see it had been disconnected 2 hours before i took the router stats. It had previously stayed connected for around 8 days before.


his ADSL has a crap load of FEC erros and error seconds for a 2hr uptime, there is an issue with his line at some point.
Are all these fec errors and error seconds bad? They might be produced from the extension cables, can't leave the router plugged in to the master as there are no power sockets near and it is also stupidly placed
 
Last edited:
Soldato
OP
Joined
14 Oct 2003
Posts
13,451
Location
South Derbyshire
I believe the line back to my exchange follows the main roads, which comes to around 0.9 miles or 1.44Km.

Tried the test port on the master socket and that only resulted in a better ping. We do connect the router to the master socket via an extension which is plugged into another extension (God knows who wired this).

As for electrical noise our line does enter a few foot from the fuse box, i don't know if that would make a difference. There is also a large transformer looking thing down by the cabinet.

As for resyncs, you can see it had been disconnected 2 hours before i took the router stats. It had previously stayed connected for around 8 days before.

Poor quality cables can result is poor stats.


Are all these fec errors and error seconds bad? They might be produced from the extension cables, can't leave the router plugged in to the master as there are no power sockets near and it is also stupidly placed
 
Back
Top Bottom