Only 800kbs on 15MB estimate ADSL2+???

Good morning.
First post, so naturally that can only mean I am after some advice! Please be gentle

We moved into a new house a few weeks ago. The house is quite rural so weren't expecting too much from the broadband speed.

We were quite surprised when we ran the speed-checker on BT's website and found that they meter our telephone number and line with a range of between 8MB and 18MB with an estimated download speed of 15MB.
The exchange is ADSL2+ enabled.

We don't like BT customer service very much, so joined Xilo/Uno for Broadband. Their own website indicated similar figures presumably because they are using the same information provided by BT.

We are in fact achieving only around 800kbs to 900kbs.

Xilo arranged for BT Openreach to visit us who reckon we are some 6.5kms line-length from the exchange. They also mentioned that there is some old aluminium cabling causing resistance.

We're now being told that because there isn't a fault with the line, BT are charging us for their visit, which was only arranged because of the huge chasm between the speed we are getting and the speed they indicated our line could achieve making us think there was a fault.

Naturally we have disputed this charge to our ISP Xilo for that reason.

I have also taken some advice from Ofcom this morning which was actually rather interesting - The guy I spoke to told me categorically that the speed range of between 8MB and 18MB that BT have metered our line at and are publicly quoting, should be achievable at or around the quoted estimated speed of 15MB because BT aren't in the habit of making such mistakes.

However, he also said that if there is some issue with the quality of the line, BT are under no obligation to change the line out or make alterations.

I'd really appreciate some advice! I'm not entirely unrealistic - We are a way from the exchange, but it does appear there are problems with BT's hardware and that the figures they quote on their website are in fact in some way accurate. We're just not achieving them for some other reason. If we could achieve even 2 or 3MBs that would be a result. 800 to 900kbs is just not workable.

Is there anyone we can talk to / anything we can do to get this resolved?

Thank you in advance

Xilo would have access to line attenuation data which should show if you are a long line.

What ADSL modem do you have? Grab the connection speed, attenuation and noise margin data from it when

a) Using it in normal location
b) Using it in the master test socket

and then use http://www.coolwebhome.co.uk/calc to get an estimate for how the line is performing.

If the telephone number was NEW to you, then it may have been showing the speeds from its previous location. What speeds do the postcode and full address check show.

First, please try to use the right units ... it will make understanding your post a lot easier.

b - bit
B - Bytes

download speeds in kbps or Mbps


Now, the first pieces of information needed to help you out are the Line Stats: Attenuation, SNR and Sync Speed for both Up and Down. You should be able to find those in your Modem/Hub.

With that an estimate of straight line distance to the exchange and the distance by road - just to see if the 6.5 kms is realistic.

The run a speedtest on this site and post a link to the results.

Edited by MHC (Fri 20-Sep-13 10:47:38)

amyw - my experience of the BT Wholesale speed estimator has been terrible. Heaven only knows which algorithms they use to estimate speeds, but they were offering speeds of 'up to 4.5mb' to premises ion my exchange who were only getting a 'poor' dial-up speed and were around 9km line length. Unreal.

An email to the then CEO LIvingston produced a rapid revision of speeds, but still 'up to 3mb' which is physically impossible as afr as I know, and a BT engineer confirmed there was no way the properties could ever receive any 'broadband' at their distance..

There are two possibilities:
1 BT lie
2) BT's calculations on longer lines are wildly inaccurate

Both give me serious concern for the results of the much-vaunted BDUK NGA scheme particularly in rural areas where long line lengths are common.

I reckon 800kb is a reasonable result at your line length. I suspect the reason they wish to 'charge you' is that there is no obligation for BT to maintain your line for anything other than voice/fax and your line does support that. I would hold out very little hope for you or the 'high speed broadband' solution from your County plan too and I suggest you start looking and agitating there.

Sorry to be depressing!

Nevermind the online availability checkers, what expected speed did ISP give you during sign-up, as they are required to do by Ofcom?

If your actual speed is significantly below this, then you are free to leave and your ISP should bear any costs in investigating the discrepancy.

BT are charging us for their visit

No, you are not in contract with BT. Your ISP might be trying to pass on BTOR's charges to them, but they shouldn't be.

the questions are.

did the isp inform you of charges prior to the visit?
how was the estimated speed relayed to you at point of sale?

I certianly would be disputing the fee like yourself although I am not sure at this moment on how I would go about it, I do think its poor for the isp to blindly pass on the charges tho.

In this case, the data was checked and BT admitted that the line data was wrong. This was relayed to the customer.

We pushed back to double check this because of the large discrepancy and because no fault could be found with the line, was offered a SFI.

We gave the customer the option of a SFI and explained about potential charging and under what situations. The text asks for time/dates for the visit and also to make a note of the engineers name, time of arrival/departure and what is done/said - so that when disputes are required, we generally have lots of information to work on.

What Openreach engineers say to end users and what they report back can often be very different, so having a log of what happened has helped in the past dispute notes that engineers leave.

Within the text, we also explain about disputes and the time frames for them to be raised. The customer in this instance has requested we raise a dispute which we have and until the outcome of this, we do not pass on the charge.

Matt

In reply to a post by MrSaffron:

line attenuation data which should show if you are a long line..

80db was what was seen from the original line data at the remote (exchange) end.

Aluminium was also found by the engineers that visited, which won't be helping.

Matt

If that was vaguely in line with distance from exchange, then booking a SFI was only every going to be a long shot and would only have improved things if able to rip out a few km's of wiring, or switch line from Al to Cu

In reply to a post by uno:

In this case, the data was checked and BT admitted that the line data was wrong. This was relayed to the customer.

I wonder if the estimate was based on stale data - where the number used to exist - rather than where it resides now after it was installed at the new location?

Hi again.
Thank you for all of your replies. Carried out the test that MrSaffron suggested with the following results:

Faceplate on:
Downstream Connection Speed: 1002 kbps
Line Attenuation: 81db
Noise Margin: 6.4db

Faceplate off: 979 kbps
Line attenuation: 81.5db
Noise Margin: 6.6db

Using the faceplate-off results on the coolweb calculator indicated that we should be getting:

ADSL2: 500kbps to 1200kbps (current result: "Slightly Low")
ADSL2+: 1200 to 2400kbps (current result "Bad")

What should my next step be?

Xilo say there is no fault on the line and that the low speeds are entirely caused by the long line to the exchange.

The results from the coolweb calculator seem to be indicating there may be some other problems.....Any help very, very gratefully received....If we could get even close to 2000kbps we would be very happy indeed!

Thank you

Edited by deleted (Fri 20-Sep-13 18:14:03)

At 81dB you are outside the documented experience of ADSL, i.e. none of the theory or feedback from around the world gives data at those sorts of attenuations.

81dB, does suggest a line of 8km length, so if that fits with the knowledge of village the exchange is in then, sorry but that is all you will get from ADSL.

Booking the SFI was a waste, unless you had very good reason to be believe the line was in reality a lot shorter.

2000 Kbps - no chance. For those speeds its find a fixed wireless provider, satellite or another alt-net, or wait for your counties BDUK project to deliver.

In reply to a post by MrSaffron:

At 81dB you are outside the documented experience of ADSL,
...
2000 Kbps - no chance.

In which case don't you need to fix your calculator at these attenuations?

It deceptively gives ranges variously as:

• 300Kbps to 2033Kbps
• 500Kbps to 2200Kbps
• 1200Kbps to 2400Kbps

EDIT: Just sussed out these are the 63dB attn. ranges, misrepresented as those for 81dB. All the more misleading!

Booking the SFI was a waste

by the ISP who should be paying for it.

Edited by XRaySpeX (Fri 20-Sep-13 22:21:51)

Yes boss

OK! Seen my edit?

I hope you see this as constructive criticism or peer review.

ok good to hear your side of things matt thanls.

I dont know what I would do in this situation but I wouldnt be happy, to me the howler by BT on the estimate.

Andrew - from which table are you getting a predicted 800kb at 8km? All the graphs I have seen show it 'fizzling out' by 6km.

In reply to a post by mikejp:

Andrew - from which table are you getting a predicted 800kb at 8km? All the graphs I have seen show it 'fizzling out' by 6km.

depends on the wire diameter. 6 km certainly isn't a limit.

From the small amount of real data on lines that long.

Record seems to be around 11km

I ask because the BTOR engineer said 9km was out of the question. None of the 'accepted' ADSL line length graphs I have seen (4) seem to agree with you. Would you like to include this in your "real data on lines that long"?

In reply to a post by mikejp:

I ask because the BTOR engineer said 9km was out of the question. None of the 'accepted' ADSL line length graphs I have seen (4) seem to agree with you. Would you like to include this in your "real data on lines that long"?

Published graphs are from calculations or line length attenuation simulators. Nobody can say 9km is out of the question credibly unless they define the wire diameter. Distance per se is not a constraint.

Yes, I understand that, but in the real world, how many 'remote' locations do you think will have anchor hawser D-side lines? Do you expect all these graphs were based on 'normal' D-Side lines or particularly 'skinny' ones? We know, of course, they specify 'good quality' D-side lines ie not Alu

I take it you have no comment on BT's estimates for the 9km line (up to 4.5mb')?

It is interesting (at last) that Counties appear to be 'waking up' from their pleasant dreams.
http://www.ipswichstar.co.uk/news/suffolk_county_cou...

In reply to a post by mikejp:

It is interesting (at last) that Counties appear to be 'waking up' from their pleasant dreams.
http://www.ipswichstar.co.uk/news/suffolk_county_cou...

Typical politician hot air. He acknowledges that BT "are a commercial company, they are not a charity" and yet expects them to do things which may not be economically viable.

BT are charging us for their visit
No, you are not in contract with BT. Your ISP might be trying to pass on BTOR's charges to them, but they shouldn't be.

A good point, well made XRay.

Ignoring the gauge of the pair. ADSL becomes dubious above 70db attenuation.
I wonder if the OP might be better served by vanilla ADSL rather than 2+ ?

Certainly I understand the losses with ADSL2+ over long lines are far greater than with ADSL2.

I think the important thing here is that the realisation of the King's New Clothes is slowly entering the public domain?

In reply to a post by mikejp:

I think the important thing here is that the realisation of the King's New Clothes is slowly entering the public domain?

I think you mean The Emperor's New Clothes, and I don't see the connection. Loads of people have access to FTTC now, BT never promised 100% coverage.

As a Royalist I'll stick with King, thanks, and it was good enough for Hans C A..

No, but we are not even talking about 100% are we - I don't think any County has 'promised' that. The question is can BT even achieve what Joe Public is expecting - which takes us back to your 'political hot air'.

In reply to a post by mikejp:

As a Royalist I'll stick with King, thanks, and it was good enough for Hans C A..

Well I thought Kejserens nye Kl�der translates to "Emperor's New Clothes", but then I don't speak Danish.

In reply to a post by mikejp:

No, but we are not even talking about 100% are we - I don't think any County has 'promised' that. The question is can BT even achieve what Joe Public is expecting - which takes us back to your 'political hot air'.

He's saying what he thinks will win him votes like any politician, yeah. But other members of "Joe Public" can see his comments for what they are and won't be so impressed.

most of the graphs are for thinner gauge than one would find on a 9km line

You get 60% more distance for a given attenuation for 22 AWG compared to 26 AWG wire for example.

http://upload.wikimedia.org/wikipedia/commons/6/65/A... is a prediction that goes well beyond 70 dB.

Edited by yarwell (Sat 21-Sep-13 12:48:42)

In reply to a post by mikejp:

Certainly I understand the losses with ADSL2+ over long lines are far greater than with ADSL2.

That's a definition problem. The losses at the same frequency are the same. The extra frequencies of ADSL2+ are lost on long lines, so appear as a bigger loss, but the loss of sub 1 MHz signal is the same.

In my experience of a >70 dB village ADSL2 does substantially better than G.dmt, as indeed it should.

Your link shows ADSL2+ running out of 'puff' at about 53db!

May I repeat the question,
"I take it you have no comment on BT's estimates for the 9km line (up to 4.5mb')? " on the 9km dial-up line?

In reply to a post by mikejp:

Your link shows ADSL2+ running out of 'puff' at about 53db!

It shows it has no advantage over ADSL2 at that point because the additional frequencies have been attenuated out. ADSL2+ is a superset and encompasses ADSL2 so a multimode router on a long line would perform at ADSL2 levels or negotiate ADSL2 in the first place. What was your point ?

May I repeat the question,
"I take it you have no comment on BT's estimates for the 9km line (up to 4.5mb')? " on the 9km dial-up line?

Not sure where your "dial up" comes from, you must be some swivel eyed campaigner or something. The OP has a 1M sync speed and would get more at a 3dB noise margin.

I have inadequate information to comment about an apocryphal claim that a speed estimator was incorrect when accessed at an unknown stage for a line in an undisclosed location on an undisclosed exchange. I am aware of 9km lines that work on ADSL.

I would say that speed estimators are more often right than wrong, by a large margin, and simply don't matter at all anyway as the modem and exchange negotiate the best deal they can. First world problem.

If the OP believes they were missold the rules are in place to get them a refund.

@ Oliver341,

Having access to a service isn't the same as having the service connected if you observe what BT say re FTTC. E.g. you might have a Huawei 96 FTTC, or even a Huawei 288, but with say 500 lines in the PCP you could be unlucky.
We also don't like the area being littered with multiple FTTC just because the sums and design were wrong.

Hence why some areas end up with two or more fttc cabs when demand is high

"you must be some swivel eyed campaigner" - aha! Do I detect an enobled Conservative here?

"Not sure where your "dial up" comes from" - let me help you - this refers to post 4 on the thread, not the OP. Not 'apocryphal'. I could post the BT wholesale 'prediction' jpg if desired, but obviously de-identified on phone number. The estimate was WILDLY incorrect as the premises is confirmed as dial-up only, and we have the very real risk that if the local Council are not keeping their wits about them, BT could conveniently declare it as >2mb and 'job done'. The OP posts another 'WILDLY inaccurate' BT speed prediction.

"What was your point ?" - merely querying "is a prediction that goes well beyond 70 dB." when the evidence you provide stops at 53? What was the point of the graph?