Is this the best I can get on Fiber?

Hello Experts..

Just signed up fiber with TalkTalk couple of days ago, it was estimated that I would get about 32-35Mb download speed.

But I'm only getting about 23Mb/24Mb, can my line handle anything better from these stats?

Up
Connection status:
Showtime
Upstream line rate (kbit/s):
1999
Downstream line rate (kbit/s):
25542
Maximum upstream rate (kbit/s):
6703
Maximum downstream rate (kbit/s):
25326
Upstream noise safety coefficient (dB):
16.8
Downstream noise safety coefficient (dB):
5.8
Upstream interleave depth:
0
Downstream interleave depth:
0
Line standard:
VDSL
Upstream line attenuation (dB):
44.5
Downstream line attenuation (dB):
19.1
Upstream output power (dBm):
1.4
Downstream output power (dBm):
10.3
Channel type:
None

Thanks for any help

Is this your first FTTC connection?

Was this a self-install, or an engineer install?

The downstream line attenuation at 19.3db is better than mine (at 21.3db), and I sync at 58mbps so, in principle, I would expect your line to be able to do better than 25mbps. There are all sorts of reason why it might not, line imbalances, differences in power masking at your cabinet and so on. However, there's one thing that is worth eliminating straight away and that is any impact caused by internal phone wiring in your house. VDSL is rather susceptible to that. Note, it doesn't matter if anything is actually plugged into any of the extension sockets, just that the wiring is there.

If you do have extension wiring, and have a modern master socket, then it's worth doing a simple test. Remove the faceplate (which isolates any extensions) and plug the router into the test socket. If it gives a substantial improvement, then it's worth looking at optimising internal wiring. If that doesn't yield and improvement, then only an engineer is likely to be able to make any improvements as it requires specialist test equipment and moving pairs and the like.

Yes 1st fiber and it was a self install.

Thanks about the comments about using the master socket test socket, will try that out. My mastersocket is in the landing with no power points and stuff so can't leave it plugged in there. But will try out the speeds over the weekend plugged into the test socket by using an extension lead to power the modem.

It's often the case that there's no power socket next to the master. As you are clearly connecting via an extension socket, then it's quite likely there will be a difference. There are all number of remedial things that can be done. Some are trivially easy. Disconnecting the ring wire to all the extensions on connection 3 often makes a large difference. Others are more involved including filtering at the master and providing a dedicated VDSL feed over cat 5 cable.

However, the first thing is to see if it works better off the test socket before making any changes (although the ring wire trick is always worth doing anyway).

+1
Ensuring the OP means the test socket, not simply the master socket.

Edited by RobertoS (Wed 10-Dec-14 23:12:59)

I was kind of wondering this myself. but as someone has said. They have a higher line attenuation but can sync at a higher speed.

I would just like to ask if the connection I'm getting is good for the line stats or should it be going faster?

3. Firmware version: Software version 4.7.5.1.83.8.173.1.6 (Type A) Last updated 14/05/14
4. Board version: BT Hub 5A
5. VDSL uptime: 8 days, 20:40:36
6. Data rate: 19872 / 56801
7. Maximum data rate: 19875 / 56005
8. Noise margin: 6.0 / 6.6
9. Line attenuation: 0.0 / 13.1
10. Signal attenuation: 0.0 / 13.1

I was getting a solid 75Mb line when I first started, but then after 4 weeks, it dropped to 53Mb. I'm litterally about 200m from the Cabinet. I've heard people say that it is distance dependent, but I'm guessing that this issue could be crosstalk?

A couple of weeks ago, it shot up to 63Mb and stayed there for about a week, but then shot back down. Can anyone shed some light as to why this may have happened? By the way am on Cabinet 26, Postcode WA7 5YN. Thanks!

Does your kit give you access to the rest of the stats? As in:-

Max:    Upstream rate = 14571 Kbps, Downstream rate
Bearer: 0, Upstream rate = 14362 Kbps, Downstream ra

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):        6.4             5.6
Attn(dB):        19.6            0.0
Pwr(dBm):        13.4            6.6
                        VDSL2 framing
                        Bearer 0
MSGc:           18              26
B:              239             237
M:              1               1
T:              64              64
R:              0               16
S:              0.1366          0.5271
L:              14056           3855
D:              1               1
I:              240             127
N:              240             254
                        Counters
                        Bearer 0
OHF:            1690064908              51509
OHFErr:         280586          2158
RS:             0               1484394
RSCorr:         0               94044022
RSUnCorr:       0               0

                        Bearer 0
HEC:            330463          0
OCD:            12823           0
LCD:            12823           0
Total Cells:    3386594154              0
Data Cells:     424058556               0
Drop Cells:     0
Bit Errors:     0               0

ES:             26332           1121
SES:            801             0
UAS:            451             441
AS:             3708201

                        Bearer 0
INP:            0.00            0.00
INPRein:        0.00            0.00
delay:          0               0
PER:            2.19            8.46
OR:             87.50           30.23
AgR:            56005.22        14392.44

Bitswap:        2742695/2742696         93099/93284

These are also useful:-

Max:    Upstream rate = 14534 Kbps, Downstream rate = 56492 Kbps
Bearer: 0, Upstream rate = 14362 Kbps, Downstream rate = 55918 Kbps

Discovery Phase (Initial) Band Plan
US: (7,32) (871,1205) (1972,2782)
DS: (33,859) (1216,1961) (2793,3970)
Medley Phase (Final) Band Plan
US: (7,32) (871,1205) (1972,2782)
DS: (33,859) (1216,1961) (2793,3970)
                  VDSL Port Details               Upstream                Downst
ream
Attainable Net Data Rate:           14534 kbps              56492 kbps
Actual Aggregate Tx Power:             6.6 dBm               13.4 dBm
=======================================================
        VDSL Band Status                U0              U1              U2
        U3              U4              D1              D2              D3
  Line Attenuation(dB):  5.4     29.7    45.1     N/A     N/A    14.4    36.8   57.1
Signal Attenuation(dB):   N/A    28.6    43.1     N/A     N/A    19.1    36.4  57.2
                SNR Margin(dB):   N/A    5.6     5.5      N/A     N/A    6.4  6.4     6.4
                 TX Power(dBm): -12.9   -13.2    6.4      N/A     N/A    10.9 7.8     5.5

Is this in relation to the original poster or my post? If it is my post then mIne doesn't. Only can get Home Hub 5 stats when I put the old 192.168.1.254 in the browser.

It was to you, as per the header link .

As you can see, the info available from an unlocked Openreach HG612 modem is considerable more useful. Tells us all about the interleaving regime, latency delays, bin range usage and so on.

They help diagnose problems.

it was estimated that I would get about 32-35Mb download speed.

- I guess TT would have taken this from the BT Wholesale checker for your property. It would be interesting to return in a few weeks to the checker and see if BT's 'estimate' has changed. Mine and several others around here have, following install, and ALL are showing worse estimated speed performance. Interesting.......

The BT Wholesale checker has always adjusted over time (on ADSLx) to the actual for the particular line.

Ok So tested in the test socket,

and got
Down line rate = 34021kb
Max Down rate = 41436kb

big difference!!

So I disconnected the extension upstairs as all the phones are connected downstairs and on the master socket got
Down line rate = 31565kb
Max Down rate = 37698kb

still fairly good.

So now back to where the modem is normally kept, after disconnecting the upstairs wires now I get.

DSL synchronization status:
Up
Connection status:
Showtime
Upstream line rate (kbit/s):
1999
Downstream line rate (kbit/s):
28205
Maximum upstream rate (kbit/s):
6454
Maximum downstream rate (kbit/s):
33497
Upstream noise safety coefficient (dB):
16.5
Downstream noise safety coefficient (dB):
6.2
Upstream interleave depth:
0
Downstream interleave depth:
0
Line standard:
VDSL
Upstream line attenuation (dB):
43.7
Downstream line attenuation (dB):
18.4
Upstream output power (dBm):
0.2
Downstream output power (dBm):
11.1
Channel type:

A 3 Mb increase on connection and 8Mb increase on the max rate.

2 questions guys,

Why is it not connecting on the max available speed?
and when I try a speed test, the results are still showing only as 24Mb.

Why is it not connecting on the max available speed?

I'm not well-versed in the lowdown of the way the TT equipment reports stats - I'm better with the quirks of the BT Huawei modem.

However, it looks like your line has had DLM intervene. The clues are that your noise margin has remained around the 6dB value, which is standard, while reporting this higher value in the attainable.

When DLM does a first attempt at intervention, it usually steals between 10% and 20% of your bandwidth, and re-uses it for error correction purposes. It also adds 8ms latency to the downstream process, which usually ends up nearer 10ms on a complete round-trip ping.

The max attainable figure is usually derived with the assumption that all that error-correction bandwidth can be given back to the standard sync speed.

BUT...

On the Huawei, the modem over-estimates this max attainable figure. If DLM de-intervenes, and returns the error-correction bandwidth back to the standard sync speed, it will normally only gain back about half of the difference.

I have no idea whether the TT modem suffers from the same kind of quirk - so you might truly be able to get to 41Mb in the test socket, if DLM relents, but it might only come out to be 38Mbps.

OVERALL...

2 targets:

1. Get DLM to relent.
DLM intervenes because it sees too many disconnects, or because your error rate is too high. If the latter is the cause, the first suspect would be your internal wiring.

2. Fix your internal wiring
Those huge differences are a sure indication that your internal wiring is bad.

Fixing this so that you get speeds close to the test-socket value is a first step. Once achieved, you then have to hope that you get fewer errors, and that DLM will de-intervene.

If you give more details about the way in which your phone sockets are wired, people will be able to offer advice.

From what I can find out, it seems that the difference between the maximum achievable rate and the actual rate is the amount of bandwidth that's used for error correction. On a noisy line, then there's more provision for correction and a lower data rate. Another possible explanation is that it's the rate that would be negotiated if the noise margin was 0dB, but I don't think that's the case as my sync and max data rates are reasonably close despite a 6dB noise margin.

A bit disappointing that the speed test throughput results haven't improved, but I don't know how TT manage their data speeds so it might or might not get faster throughput with the slightly higher sync speed.

As to improving your current situation, you've made a good start in disconnecting unused extensions. However, would you ever want to reconnect them, I'd strongly recommends getting a VDSL Interstitial Faceplate as that provides the means of filtering all the phone extensions plus a connection for an unfiltered VDSL extension.

http://www.amazon.co.uk/GENUINE-BT-Openreach-Interst...

To try and get your VDSL extension to get close to your test socket performance, then the first thing to do is disconnect the ring wire at the master socket end (doing it at the other end won't work). It's essentially a "free" upgrade and the ring wire is essentially redundant with modern phones (and totally so when you use microfilters).

However, if that's done, then the next step is more work, and that requires using twisted pair cabling for the phone extension. The twisted pair makes for a better balanced line, and is much less vulnerable to interference. The "gold standard" for this is cat5/6 ethernet cabling. That has 8 wires connected as four pairs. It's readily available and not expensive, but the downside is that it's rather fatter than cheap phone extension cabling. I use this type of cable and I feed both a filtered and unfiltered feeds on separate pairs to an extension socket which has connections for both phone and router.

Over about 7 metres of cat5 I get virtually no degradatio of signal.

A second best choice is to use BT spec 4 core CW1308 extension cable which is considerably thinner than ethernet cable. (There are versions with more cores). The 4 cores are comprised of 2 pairs. You can then use one pair for an unfiltered feed and use a microfilter as I assume you do. It's about the same standard cable as is used in the main network, so I think it's probably good enough.

It's widely available (Maplins sell what they say is CW1308, but it doesn't mention twisted pair on the website, so I'd check). This is an alternative supplier

http://www.run-it-direct.co.uk/CW1308cable10MW.html

Thanks Guys,

I had already removed the ring wire back in the ADSL days..

I probably have some DLM issues at the moment as I have been connecting and reconnecting at different sockets and test sockets in the last day.

After removing the upstairs wiring I saw a change in sync speed, this got me interested. I can't really change cables as the wiring is really neatly done by the floorboard and don't want to make it all look messy.
Instead though I will reduce the amount of hard extensions used, so removed the extension to the kitchen and the results are surprisingly good!

The new stats after removing all but the single extension to the modem and cordless.

DSL synchronization status:
Up
Connection status:
Showtime
Upstream line rate (kbit/s):
1999
Downstream line rate (kbit/s):
34357
Maximum upstream rate (kbit/s):
6842
Maximum downstream rate (kbit/s):
40394
Upstream noise safety coefficient (dB):
16.3
Downstream noise safety coefficient (dB):
5.7
Upstream interleave depth:
0
Downstream interleave depth:
0
Line standard:
VDSL
Upstream line attenuation (dB):
39.7
Downstream line attenuation (dB):
18.3
Upstream output power (dBm):
-3.2
Downstream output power (dBm):
11.2
Channel type:
None
DSL up-time:
0 days 3 hours 34 minutes 17 seconds

Thats almost the same as the test socket!!

BUT! The speed tests are still showing as only 24Mbps! I'm guessing as I must have switched on and off the modem at-least 10 times in the last 24 hours testing all these wiring options have driven the DML/error correction mad and is limiting my speed?

Any other suggestion guys?

Also any idea on how long is the DLM going to take to normalise after all these restarts by me?

Edited by deleted (Sat 13-Dec-14 16:43:22)

Cheers Rob!

Will try to get my hands on one of them and see if I can pluck the stats from it. With me having a HH5, would it still be fine seeing as though the HH5 has its' own built in VDSL modem?

I do have a spare HH3, that I suppose I could hook up if that would be better with the HG612 modem?1

Also does the DLM always stay on in the FTTC cabinet? And does it go back to normal after couple of weeks?

DLM intervenes generally the most in the first day or two but remains active forever, and can go faster if things improve.

DLM limits the connection speed and the level of interleaving, so should not be the reason why if say connecting at 34 Mbps you are getting 10 Mbps slower than that. With 34 Mbps connection maximum speed test should be around 32 Mbps, but assumes a perfect connection and computer setup. Our speed test is sensitive and shows enough information you can tell the difference between a perfect, and something is messing things up.

distance still plays a part on VDSL, but crosstalk also plays a part, and crosstalk can play a massive part.

I have seen lines with 50% higher attenuation than me syncing higher than me, but I have also seen lines with lower attenuation syncing below spec as well, its luck of the draw.

Vectoring cannot come soon enough, that should make things a bit more predictable like ADSL was.

DLM certainly can reduce intervention, and remove it entirely. It may take a couple of days... and it might take a couple of months. It really depends on the amount of errors being seen on your line.

As MrS says, your download speed in test should only be a couple of Mbps lower than the sync speed, so you should be seeing around 31-32Mbps.

One reason: are you running the test over WiFi or a powerline converter? Those could cause a bottleneck; using a direct ethernet connection is the best way to test.

In reply to a post by WWWombat:

DLM certainly can reduce intervention, and remove it entirely. It may take a couple of days... and it might take a couple of months. It really depends on the amount of errors being seen on your line.

Indeed. My line used to get interleaving put on it now and again and it would be removed 14 days later in the early hours. Thankfully there was a firmware update last year and since then I've syncd higher and been completely stable.

It seems it was something to do with a limit on TalkTalk end, so I switched off the router for 20 mins and back on again.

The sync was still at 34Mb, but now the download tests are showing speeds of 32.5Mb!!

I'll wait for the DLM to get the sync to get closer to 39999 and switching off the modem for a while for Talktalk to resent their speed limits at their end.