From 512k to fibre - your biggest increase on long lines?

Hi,

I'm curious to see how many people on FTTC have achieved a dramatic improvement/uplift. My mum's broadband is about 750k in Devon, and her exchange will soon be fibre-enabled. The cabinet is far, far closer to her house than the exchange (which is 4km+ away), so her uplift could be 50 times greater in theory, or more.

Have any others here seen such dramatic changes? Long line to the exchange, but a much shorter run to the cabinet and therefore a revelation in speeds?

Would be fun to hear from rural places like her's, and see how fibre is transforming areas.

Will

I've gone from 2meg to 53meg (although now its down to 32) and a friend has gone from 1.5 to 75meg, both of us live in Devon and our exchange is in another village (estimated 5km away) but our cabinet was in ours and so obviously much closer.

I could give you some long line figures but seems you only interested in rural.

MY data is from sub 1mbit urban lines, the only rural lines I know off are sub 10db attenuation, to me the rural only long line thing is a myth, they in rural and urban locations.

Brilliant. Whereabouts in Devon?

True, I'm biased towards rural! But there's a reason for rural locations being the last "connected" area of the UK, and that's the line length and topography. Of course long lines exist everywhere, but far fewer in urban areas which benefit from greater population and therefore demand for investment.

The existing ADSLx speed is completely irrelevant as it depends on distance from the exchange. FTTC speeds depend entirely on the copper distance between the FTTC cabinet and the premises.

Estimates for most lines on the Openreach commercial rollout are available from this checker, using either the phone number or address. The postcode checker option is not to be trusted.

That checker should give you the cabinet number the line is connected to, even if there is no FTTC estimate. Given the cabinet number you need to find it if there is no estimate.

not completely irrelevant.

large distance on E side means more agressive adsl power cutback.

Exminster

My parents are on a 1mb connection, and Exchange enabled for fibre. Once their cabinet is enabled (no ETA) they should get 45mb

The effect of the power cutback does change with (E-side) distance, but with a complicated interaction of the number of affected frequencies vs the actual power reduction for each frequency. So the longest E-sides get the highest power reductions, but over the smallest set of frequencies.

Judging from the tests specified by BT (for ISP-supplied modems) it appears to be the cabinets with medium-length E-sides that suffer most, longest-length E-sides suffer least with the shorter-length E-sides inbetween.

http://forums.thinkbroadband.com/fibre/4317708-bt-si...

Robertos is correct regarding ADSL not being a good indicator because of the very significant differences in Line Length.

However if you use the BT Checker he correctly refers you to, take particular note of the second and third paragraphs-

For FTTC Ranges A and B, the term "Clean" ...

Throughput ...


Although referring to FTTC, those aspects also apply to ADSL and to normal Voice Phone usage.

So if you have not already done so, you should use the Quiet Line Test to ensure your mother's Land Line is already in good condition, including within her house, taking appropriate remedial action, if required.

If such faults are between the cabinet and her house and/or are internal to her house, they have proportionately much worse effects on FTTC/VDSL, because of the higher frequencies involved.

I'm in the village of Bramley which had historically suffered from being over 4.5km from THTG (Turgis Green) exchange, a mix of aluminium/copper, over head wires and more rabbit runs that Watership Down on viagra.

Back in the ADSL days, I'd get 2Mbps down, 250kbps up.
Thanks to a nice shiny new green cab less than 150m away, I'm now getting over 60Mbps down, 18Mbps up and it works flawlessly providing the line doesn't break (we've had 2 breakages in the same place within the last few months and my DLM is still recovering from a copper break two weeks ago but all going well so far).

ADSL Downstream Rate 1.5Mbps / Downstream Range 1 to 3.5Mbps

FTTC Clean: 80 / 72.2 20 / 19.9
FTTC Impacted: 80 / 60 20 / 17.9

I'm on Plusnet with the pro-addon (more for reasons of paranoia rather than necessity) and average around 250Gb of traffic a month. Have 3 kids between 15 and 19, who stream music, videos and general usage as well as IPTV.
The ability to stream HD IPTV channels, watch Netflix, stream music combined with my working online has been flawless. No buffering, queuing, timeouts and it doesn't matter how hard we appear to hammer the connection, it works amazingly well.

The stats in my signature were taken when I was the 3rd live connection in my cabinet. I believe my cabinet is up to around 30 connections. The wiring between me and the cabinet is a mix of aluminium for the 1st 100m, then into a 100 pair not so water tight junction box in a chamber that constantly floods, from where there is then a 50m run over overhead lines also carrying mains power to the house.

Spookily, the joining of aluminum and copper combined with being underwater confirms that Lougi was spot on. In 10 years, that joint has failed 5 times.
Continues to be the only weak point in my system at present.

I have analysed dozens of snr and bitloading graphs.

The general pattern is longer E sides lose a lot more sync speed to power cutback than shorter E sides, its not equal distribution as you suggest.

I have seen short E sides have barely any cutback at all to the point its hard to notice it unless you looking for it, they lose 2-3 bits and thats it.
I have also seen long E sides where the power cutback has removed over 100bits of bandwidth. On my line I have some tones down to almost 0 bitloading on the adsl frequencies.

The reason is short lines on adsl even at the top end of adsl2+ range still have a much stronger signal than long lines at the low/medium frequencies. Its also possible BT have gone over the top and are over compensating, but regardless of the reasons, the pattern is there, long E-sides lose more bandwidth than short E-sides. Not to mention power cutback is on lower frequencies which without power cutback would get higher bitloading than the higher adsl2+ frequencies. So even if the cutback was equal, the longer E-sides would still have a higher penalty.

my adsl was 50db attenuation, the power cutback is affecting from 0 to 260 tones, that seems narrow to you?

Edited by Chrysalis (Thu 12-Jun-14 09:02:58)

In reply to a post by eckiedoo:

Robertos is correct regarding ADSL not being a good indicator because of the very significant differences in Line Length.

[...]

So if you have not already done so, you should use the Quiet Line Test to ensure your mother's Land Line is already in good condition, including within her house, taking appropriate remedial action, if required.

If such faults are between the cabinet and her house and/or are internal to her house, they have proportionately much worse effects on FTTC/VDSL, because of the higher frequencies involved.

Understood. Thanks for the clarification. I suppose my point (particularly for her setup) is that we've done all we can in the house; the socket and wiring is fine, and the line tests are always clean (audibly so). The line attenuation and stats are truly horrific, almost off the scale - but that said, the connection is rock solid stable at 900-1000Kb/s.

The reason it's so poor for most of the village is predominantly down to line length (5+ KM), so bringing the DSLAM into a shiny new cabinet ought to (!) improve things dramatically. Instead of a 5km line length, it's effectively under 1000m for the entire village. I think I estimated the cable run from her house to the cabinet ought to be about 800m.

(existing stats below)

BT Test xDSL Status Check:Pass OK.pass OK. Circuit In Sync
BRAS=750Kb/s FTR=921Kb/s MSR=1152Kb/s ServOpt=1 I/L=A
Up Sync=448Kb/s LoopLoss=40.5dB SNR=15.3dB ErrSec=0 HECErr=0 Cells=1131
Down Sync=992Kb/s LoopLoss=63.5dB SNR=7dB ErrSec=0 HECErr=0 Cells=1141

In reply to a post by Chrysalis:

The general pattern is longer E sides lose a lot more sync speed to power cutback than shorter E sides, its not equal distribution as you suggest.

What's an E side? Line length?

In reply to a post by shinerweb:

Thanks to a nice shiny new green cab less than 150m away, I'm now getting over 60Mbps down, 18Mbps up

You should be getting maximum 80Mbps sync at that distance.

In reply to a post by shinerweb:

I'm on Plusnet with the pro-addon (more for reasons of paranoia rather than necessity)

So you have £5 a month to throw away? For 99.9% of users it is a waste of money, try it without and if you notice any adverse effects you can always put it back.

In reply to a post by shinerweb:

I'm in the village of Bramley which had historically suffered from being over 4.5km from THTG (Turgis Green) exchange, a mix of aluminium/copper, over head wires and more rabbit runs that Watership Down on viagra.

Back in the ADSL days, I'd get 2Mbps down, 250kbps up.
Thanks to a nice shiny new green cab less than 150m away, I'm now getting over 60Mbps down, 18Mbps up and it works flawlessly providing the line doesn't break (we've had 2 breakages in the same place within the last few months and my DLM is still recovering from a copper break two weeks ago but all going well so far).

Dramatic difference then. Do you have your old line stats frmo ADSL days?

E side is the cabling from the Exchange to the PCP cab, D side is from the cab to the home etc. (well not quite but that's the simple way to look at it)

Edited by deleted (Thu 12-Jun-14 10:03:57)

Good, so your mother's system is well-prepared to get maximum benefit from FTTC/VDSL.

In reply to a post by deiwise:

I'm curious to see how many people on FTTC have achieved a dramatic improvement/uplift. My mum's broadband is about 750k in Devon, and her exchange will soon be fibre-enabled.

Going back a few years I was rejected for the original 512Kbps ADSL offering circa 2001. Eventually got ADSL and reached 3.5Mbps with both Sky and O2 LLU connection.

I'm now on BT Infinity and can get the full 80/20 speeds. Check my speed tests in the signature - occassionally I get more than 80Mbps

In reply to a post by Squirrel:

I'm now on BT Infinity and can get the full 80/20 speeds. Check my speed tests in the signature - occassionally I get more than 80Mbps

Sadly you don't. Bad measurement by the speed tester so best not to raise the poster's hopes.

Some people on this suburban estate went from sub-1Mb to full 76Mb.

We are actually above the EU average for our rural areas, it's our urban and suburban areas that are below the EU average bizarrely.

In reply to a post by Ignitionnet:

Sadly you don't. Bad measurement by the speed tester so best not to raise the poster's hopes.

I was thinking that. The Speedtest.net site can sometimes seem to 'overrun' at the higher speeds.

Also the flash player causes no end of grief, just see MrSaffron's thread today in TTTS.

Didn't know that. Thanks. Interesting.

In reply to a post by Ignitionnet:

In reply to a post by Squirrel:

I'm now on BT Infinity and can get the full 80/20 speeds. Check my speed tests in the signature - occassionally I get more than 80Mbps

Sadly you don't. Bad measurement by the speed tester so best not to raise the poster's hopes.

Sadly for you I do, and not only measured by speedtest.net. They just happen to be some tests I ran years ago. I can drive several file downloads using upto 50 threads per server and am virtually guaranteed to get the full 80Mbps, occasionally seeing aggregate speeds of 82/83/84Mbps for short periods.

In fact here is one I prepared for you a moment ago - https://dl.dropboxusercontent.com/u/72454955/Capture...

No, you don't. There's probably an antivirus that's buffering the files and causing the incorrect results.

You can't download above 77.44Mb, this being your IP profile on the BT Wholesale network

Still keep telling yourself you are if it makes you feel good. You are unique on the entire FTTC network in that you can download faster than your line runs, and indeed the entire world in that you can download faster than your DSL line runs without using compression, so well done and congratulations!

Edited by deleted (Thu 12-Jun-14 21:29:08)

As ign said you can't get faster than the BTO's FTTC line profile which is 77.4

which including overhead is why its advertised as "upto" 76

But speedtest is ok for testing quickly

i'd rather use TBB's speedtest

you can click the link on my sign below that is a max synced line 79999/20000 is an example of "realworld" speeds.

you can see the burst TBB just flatlines.

its not like we are trying to flame or start agro but it is physically impossible considering how the network functions.

That dropbox is without a doubt giving a false reading though it could be something else causing it (anti virus ect) i've seen myself getting over 100mbps due to some anti viruses when i only had 40/10 package.

Edited by epyon (Fri 13-Jun-14 02:55:01)

Some old stats from March 2013

Text
1 23 45 67 89 1011 1213 14	25/03/2013 07:49  DSL Upstream Rate:   448 Kbps DSL Downstream Rate: 3936 Kbps                        Down         up   DSL Noise Margin:     6.3 dB      17.0 dB DSL Attenuation:     54.0 dB      31.5 dBDSL Transmit Power:  18.9 dBm     11.9 dBm  gateway pcl-ag06.  25/03/2013 07:531720 kbps (215kB/s)194 kbps (24.3kB/s) 25/03/2013 07:521348 kbps (169kB/s)193 kbps (24.1kB/s)

In reply to a post by kasg:

In reply to a post by shinerweb:

Thanks to a nice shiny new green cab less than 150m away, I'm now getting over 60Mbps down, 18Mbps up

You should be getting maximum 80Mbps sync at that distance.

In reply to a post by shinerweb:

I'm on Plusnet with the pro-addon (more for reasons of paranoia rather than necessity)

So you have £5 a month to throw away? For 99.9% of users it is a waste of money, try it without and if you notice any adverse effects you can always put it back.

Whilst the first drop from my house to the underground junction box via over head poles has recently been replaced, the 100pair cable from the junction box to the cabinet performs slightly better than damp string.
Getting BT to even look at that box when you are between the 60 --> 80 Meg expected range is nigh on impossible.
Whilst it would be nice to get the max and given what I used to get combined with the fact no matter how hard I push it, it works flawlessly, I'm not going to quibble over a missing 10-16Mbps just yet...

Right, you've persuaded me to give dropping Pro a go...