Print

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

[tallp]

It's not unknown for the line to take a very roundabout route, so what the ISP has told you isn't impossible. But it's only a figure in a database, and even if it's wrong it doesn't affect actual performance. I think that the only way to be (fairly) certain is to subscribe to a BT-based ADSL service and get the connection stats from the router.

If I understand you correctly, you're currently with a cable provider and you want to change to a BT-based service. Depending on the ISP you *may* be able to get the line length checked some other way, but you would probably need to go with one of the more expensive, more professional ISPs such as Zen.

I can understand a roundabout route, but that would cover half of Bristol! I cant really afford an expensive line investigation to be told that some "muppet" has entered the wrong figure in the database. BT stated that I was only 600m from the exchange but more importantly I think that I should get upto 6Mb download speed. the ISP are stating that they will not even connect me with that [5780m] distance as it is outside of their service deliverables.

I dont know who to believe; chance it and go with the ISP and be stuck with no service or crawl back to my Cable company?

oh the uninformed decision we have to make!

Thanks for your speedy response btw.

Cheers

Paul   

[orainsear]

That seems an awfully long line length in comparison to your geographical distance from the exchange.  Do you have any neighbours who use ADSL that you could possibly speak to?

Which ISP was it that you spoke to?  Have you tried any others?

[tuftedduck]

Paul, go to the Kitz adsl checker page at  http://www.kitz.co.uk/adsl/adslchecker.php , enter your phone number and postcode in the appropriate boxes and you will get all sorts of info including your actual line length.

If nothing else it will give you a reliable figure on which to base any further debates with BT/ISP

 

[roseway]

you will get all sorts of info including your actual line length.

No, sorry TD, but the Kitz checker, like all other publicly accessible checkers, only provides geographical distances, not actual line lengths.

[tuftedduck]

 

[HPsauce]

I thought it accessed the BT database.

[orainsear]

It uses the Samknows database API.

Edited: Oops my bad!

[kitz]

The approx line length is based on geographic figures on the assumption that most of BT's cabling tends to follow main roads.. I use google maps and a rather complicated algorithm to get the figures.
Another slightly better indication is if you have any line stats and taking the figure from your attenuation.

The estimated speed is taken from the BTw database, and will likely be the same results as all other ISPs/checkers.

One thing that does occur.. you mention coming over from cable..  in which case is this a brand new telephone line?

If so then its likely that the figures are based on the location of whom ever had that phone no before you.
It can take a while before the BTw database update..
Some ISPs are more understanding than others if you mention that it is a brand new phone line, and highly likely that the BTw database hasnt yet been updated.

[waltergmw]

If you have a modem compatible with DMT tools there is sometimes an electical line length estimate on the right hand side written in grey in the third panel down before you start the noise monitoring. The picture below is from a netgear DG834V4 (Broadcom chip set).
I do not know how accurate it is but this value was similar to the figures provided by Zen on this line.

If you happen to live in Surrey the SCC provide quite a good interactive map at:-

http://www.surreycc.gov.uk/sccwebsite/sccwspages.nsf/LookupWebPagesByTITLE_RTF/Surrey+Interactive+Map#

and then click on the green writing. The suite contains a road distance measuring tool but again this will only be an estimate as it's almost impossible to know the exact cable route.

Kind regards,
Walter

[attachment deleted by admin]

[kitz]

>> line length estimate on the right hand side written in grey in the third panel.
>> I do not know how accurate it is but this value was similar to the figures provided by Zen on this line.

It's a broad estimate based on the attenuation (loop loss).

My maximum speed calculator tool based on Consultels survey for adsl loop loss, a graph of which can be seen here.  It involved much plotting of figures and adjustments for UK gague cable and provides an estimations for line length v attenuation.  Obviously it cant allow for such things as aluminium joints etc, but it is generally recognised as being one of the most accurate without physically being able measure it from the EU premises. 
Leo may confirm for sure,  but I think Zen (and a few other ISPs) use it if they cant get a  BT measure.

>> The suite contains a road distance measuring tool

Nice tool Notice its done by the Ordnance Survey who also approve and use the  Vincenty inverse formula for geodistic calculations. 
Quick note..  from what I saw it uses straight line between 2 points which is the calculation often provided by most sites.
Driving distance as per the road and curvatures is a bit more complicated..(hence the combo of the google maps and the inverse Vinventy formula which was a PITA to code) and afaik this site was the first and only (aside from one that copied my code) adsl site that does so.
Big thanks and credit must be given to Sam though, cause I pull on his data for the exchange info and location.

[waltergmw]

@ Kitx

>>>>>> from what I saw it uses straight line between 2 points ............

That is correct but if you have the patience (you'll certainly need it for accurate measurements !) and zoom in to a reasonable resolution on a large screen you can join as many straight lines as you like.

Kind regards,
Walter

[Azzaka]

What is the ISP Paul.

At the end of the day I have provided a line where no other ISP would and fought to keep it. As they have all stated thus far, the line distance is an estimate and the BTW database does take awhile to get updated.


Any more modem info, stats wise, would certainly help us help you.

Kindest Regards,

Leo

[asbokid]

If you have a modem compatible with DMT tools there is sometimes an electrical line length estimate on the right hand side written in grey in the third panel down before you start the noise monitoring. The picture below is from a Netgear DG834V4 (Broadcom chip set).

I do not know how accurate it is but this value was similar to the figures provided by Zen on this line.

Hi... I managed to get DMT-UX, the DMT port for Unix [1] working with the Huawei HG612, the OpenReach VDSL2 modem. Once it's more stable, I will upload the patches to some webspace.

The source code for the function that calculates line length is replicated below. It doesn't look very complicated.

dsl_chancharlog.data is the Hlog value for each tone, as parsed from the output of xdslcmd info --Hlog.

But what does the Hlog metric actually capture? And why does the function pluck out tone number 231? What's special about that tone?

Code: [Select]

char* linelength() {
int tone231;
int linelength_min;
int linelength_max;

char* text;
tone231=(int)dsl_chancharlog.data[231];
if (tone231>0) {
//search minimum chancharlog value
unsigned int i;
int chancharlogmin;
chancharlogmin=100000;
for (i=DROP_TONES; i < dsl_chancharlog.last; i++) {
if ((int)dsl_chancharlog.data[i]>0) {
if ((int)dsl_chancharlog.data[i]<chancharlogmin) {
chancharlogmin=(int)dsl_chancharlog.data[i];
}
}
}
//(MD_CHANCHARLOG_TABLE[232]*1000)/20.5 *((low_Misc_Value - 50)+100)/100)
linelength_min= (int)(tone231 * 10/20.5 * (chancharlogmin + 50));
//(MD_CHANCHARLOG_TABLE[232]*1000)/13.4)
linelength_max= (int)(tone231 * 1000/13.4);

char* text;
text=dmalloc(sizeof(char)*(8 + log(linelength_min) + log(linelength_max)));
sprintf(text, "%im - %im",(int)linelength_min,(int)linelength_max);
return text;
} else {
text=dmalloc(sizeof(char)*(26));
sprintf(text, "couldn't estimate line length");
return text;
}
}
The function returns a line length of 239m - 428m for our line which sounds about right.

It would be good to port this version of the DMT tool (back) to Windows, if anyone fancies a challenge?

Naturally, it would remain Open Source for others' benefit.  I would anticipate building it under minGW [2].
In Windows, the major hurdle would probably be the RRDtool data logging and graphing library used in DMT-UX.[3]

Are there any interested parties up for doing that?

Cheers, a

[1] http://www.spida.net/projects/software/dmt-ux/index.en.html
[2] http://www.mingw.org/
[3] http://oss.oetiker.ch/rrdtool/index.en.html

[trantore]

Hi... I managed to get DMT-UX, the DMT port for Unix [1] working with the Huawei HG612, the OpenReach VDSL2 modem. Once it's more stable, I will upload the patches to some webspace.

Hi Asbokid,

I'm also interested in refactoring the DMT-UX tool!

My suggestion is port all the code to the Qt framework, to make it platform-independent. This simplifies the interface, the socket control and other aspects of the program. Additionally simplifies the compile environment. Also, to make more simple the port to Windows, you don't need to exclude the RRDtool library, you can link to some of the two native versions:

http://www.gknw.net/mirror/rrdtool/
http://oss.oetiker.ch/rrdtool/pub/?M=D

My interest is for support more modem/routers... because the original DMT tool is close-sourced, and also the good OrbMT. So, I prefer to focus on add support to other devices.

What you think about it? When you publish the patched DMT-UX version?
Regards.

[Bald_Eagle1]

@ kitz,

Hi Kitz,

We have not spoken before, & I don't wish to hijack someone else's thread, so just a very quick query:-

Do you have any current proposals to introduce a similar checker for FTTC connections?

We all seem to be struggling to understand what SHOULD be achievable for given line lengths etc.

Paul.