Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[GigabitEthernet]

How's the line now?

[smucat]

Smucat is smiling.    FTTC Infinity 2 upgraded from ADSL 8M today. Huawei cab, ECI 1B modem and Toady HH3.0 type B.
Approx 400 metres from cab, as the cat crawls. Download speed test varying from 25 Mbps to 47 Mbps. Upload between 6 and 8 Mbps. Early days, but seems great.

[NewtronStar]

Smucat is smiling.    FTTC Infinity 2 upgraded from ADSL 8M today. Huawei cab, ECI 1B modem and Toady HH3.0 type B.
Approx 400 metres from cab, as the cat crawls. Download speed test varying from 25 Mbps to 47 Mbps. Upload between 6 and 8 Mbps. Early days, but seems great.

the best way to find Max FTTC speed is to use the estimated guide first https://www.dslchecker.bt.com/pls/adsl/adslchecker.welcome (BT) and use this as the Base and if it's way off the estimated guide 4 - 5 megs difference is fine but 10-20 megs then it sure to a line fault or internal wireing issues.

[smucat]

Thanks for the link, NewtronStar. It even gives the cab number, although not all cabs have visible numbering. Quote is 57 Mbps, but max on day 2 is still 47 Mbps. Linebox is on end of dropwire. I don't think anyone will want to replace the wiring to the cab, for the sake of 10 Mbps!!

[burakkucat]

Download speed test varying from 25 Mbps to 47 Mbps. Upload between 6 and 8 Mbps.

I am surprised that you are seeing a variation of up to 22 Mbps in the DS. I presume (as you have an ECI active CPE) that the figures you quote are throughput speed-tests? Perhaps you might consider, after leaving things for a fortnight (say) to 'settle down', swapping the ECI B-FOCuS modem for a Huawei HG612 and then using Bald_Eagle1's utility to harvest data and produce graphs, etc?

[waltergmw]

@ NewtronStar et al,

Whilst the Wholesale DSL Checker often provides reasonable estimates on shorter lines we have found, by bitter experience on many longer D side lines - this time around 2.5 km, that the enthusiastic marketing mechanism can be wildly optimistic. One such line from a cluster of 10 all suggest a preposterous figure of 19 Mbps but reality dawns as you'll see below. After two months of investigation without progress, the checker has woken up saying that line only might provide "Up to 2.2 Mbps down & 0.4 Mbps up", leaving all the others still on 19 Mbps.

Just imagine what I said when I discovered that BT sales staff, in their usual splendid "twinkling meteors" style, had cold-called the hapless family and had actually convinced them to change a second line** to Infinity !!!!
VERY fortunately I arrived before they had attempted to get a MAC.
** That second line on ADSL obtains a sync rate of 1.33 Mbps.

Kind regards,
Walter

[Black Sheep]

To clarify how the 'Checker/Estimator' operates. It only gives the approximate speed based on distance to the DP (The main pole, or underground joint if you like). The premises could then be fed via any number of 'Carrier Poles' or long underground cable runs.

As always, folk tend to pick up on the extremely rare cases BT's estimate get it wrong, for all intents and purposes, and from genuine first hand experience on a daily basis, the EU gets more than the estimate given.

[waltergmw]

HI BS and thanks for the explanation.

I know of two very poor performers from the DP which I'd guess is around 1.75 km D side distance.
Do you know what triggers the change in estimate to more realistic figures after umpteen SFIs and why all the others fed from the distant DP are not adjusted at all ?

Kind regards,
Walter

[NewtronStar]

@ NewtronStar et al,

Whilst the Wholesale DSL Checker often provides reasonable estimates on shorter lines we have found, by bitter experience on many longer D side lines - this time around 2.5 km,

Thanks for your findings, it always was my thinking that FTTC should not be provided to customers with copper wire exceeding 1500 meters (1.5KM) 

http://www.thinkbroadband.com/guide/fibre-broadband.html

unless its the faster Broadband service sub 15mbps.

[waltergmw]

@ NewtronStar,

If you were one of the unfortunates on sub 1 Mbps ADSL, experience shows that the poor blighters are relieved just to see any improvement.

@ BS I agree for all those around the 1 km D side length unless they have suffered from sub-standard subcontract VDSL installations. Those above 1 km are more variable which is perhaps understandable. The one I quoted above has a sync speed of 1.88 Mbps and is now estimated at "Up to 2.2" Mbps. Some of the longer lines with multiple DPs have higher estimates as the estimator (Only sometimes) doesn't seem to know about the unexpected drop after a particular joint feeding a DP.

We have also noted that on noisy lines even down to 1 km there is often a drop of anything up to 5 Mbps after a few days once DLM has applied interleaving. The advantage is probably that with less retransmissions and more FEC corrections, the actual throughput is more stable and sometimes just a bit faster.

Kind regards,
Walter

[bbnovice]

I don't wish to open old wounds (ie old posts I've made about this before on other threads and got my ears burned), but my experience with  BT's speed estimates has been uniformly negative.

Using the BT DSL checker mentioned in previous posts, I should expect circa 55DS/18US. This was also the estimate provided by BT as part of their marketing puff to entice me to upgrade from the old Infinity service to the (then) newly released 80/20 Infinity option.

In practice my line actually achieved 48DS/9US when initially upgraded. Recently it has averaged 47/11 - the DS has decreased but the US has improved. However - and this is very important in my case - the connection is solid (touch wood!!!) and the speed consistent. So I'm quite willing to accept the situation and have not made an issue of it with BT (particularly as the upgrade was free).

But if I had signed up as a new 80/20 customer I would not be very impressed as I can almost see the FTTC as I type this. I've never understood why on such a short run between the FTTC and my house the speed is so slow. (BTW: its nothing to do with wiring at my premises - I had a retired BT engineer in to check/fix the internal wiring after experiencing disconnection and interference problems in the old ADSL days with Virgin as the ISP - and which eventually turned out to to be a BT network problem anyway).

So from my experience I regard BT's speed estimates with some scepticism.   
 

[waltergmw]

Hi BBN,

It might just be worth checking the latest BT Wholesale checker which confirms which PCP you are connected to.
Much will depend upon how your line is actually routed.
It certainly sounds as if you are several hundreds of metres away from the actual PCP and FTTC.

I know of a group of older houses which were connected to one of the very first PCPs in Cranleigh.
Since then there has been substantial infill and TWO more PCPs added en route BUT the original (now noisy) wiring has been left in situ.
In this case the D side length is over a km bypassing the two newer PCPs and consequently Superfast sync speed only just achieves a miserable 26 Mbps downstream.

In my experience I've only ever managed to order one new line where BT Openreach were prepared to change the original wiring to a more appropriate PCP.

Kind regards,
Walter

[Black Sheep]

The old saying, 'Damned if they don't, damned if they do', springs to mind ........ in a friendly way of course.
Would EU's rather there was no estimate whatsoever ?? Also, my experience in these matters allow me to make the statement I made above ....... most EU's will achieve more speed than the 'Guesstimator' dictates, but above all, it is just an ESTIMATE, and it doesn't hide the fact that it is.

The eMLC (Managed Line Characteristics) uses Network Records and 3 other systems to calculate the line loss to THE DP. This, combined with the Cabinet CAL value (Cabinet Assigned Loss), and a model of future Crosstalk gives an indication of the long-term line rate estimation.
There are unknowns in this, such as the final drop which can sometimes be very long, the ACTUAL level of Crosstalk, or the presence of any line fault that may be there ?

Further to Walts question above, work is underway to see if development of the eMLC can be achieved to use actual measured losses of existing FTTC lines.
 

[waltergmw]

Thanks BS,

If I may second BS's thoughts on the estimator; it does state quite clearly that line speeds may drop. We have observed this too; either due to noisy lines applying interleaving at various levels or due to cross talk when more twisted pairs in the same cable obtain a VDSL service, the power is reduced so as to maintain reasonable speeds for everyone including those who have retained their older and weaker ADSL services. Another reason for lower speeds can be the banded profile purchased. E.g. Zen's cheapest VDSL service is limited to 2 Mbps upload even if the line is capable of higher speeds.

The estimator is helpful for those who would like to order the fastest possible service but would be wasting their money if the estimator is only offering say 30 or 40 Mbps. I expect in most places a small increase over the estimated speed just wouldn't be worth it - especially as the estimator is, I believe, providing sync speeds and not throughput speeds

I have one small improvement in the estimator I would like to see. That is a clear statement that an EU's cabinet has been enabled but the quality and / or length of the line prevents BT Openreach from offering any VDSL service. Currently the estimator doesn't mention VDSL at all in these cases. This idea would assist the EU as at least they would know immediately that another solution must be found if existing services are inadequate. We have one such notorious line where around 40 EUs are NOT offered any VDSL service, even though the cabinet is enabled.

Kind regards,
Walter