Print

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

[jeffbb]

Hi
I may be wrong but your DS power may be   low depending which profile you are on ?
see the some profiles below
Selection Of Vdsl2 profiles
Description                       8b       12b       17a       30a

Bandwidth (MHz)                8.83    12.0      17.66     30.0

Carriers                            4,048    2,783    4,096     3,479

Carrier Bandwidth (KHz)       4.3125  4.3125   4.3125   8.625

Transmission Power(dBm)    +20.5   +14.5   +14.5   +14.5  **

Bandwidth (Mbps,bidirectional)   50       68       100       200

** DS  power

Regards Jeff
 

[Bald_Eagle1]

Hi jeffbb,

Hi
I may be wrong but your DS power may be   low depending which profile you are on ?
see the some profiles below
Selection Of Vdsl2 profiles
Description                       8b       12b       17a       30a

Bandwidth (MHz)                8.83    12.0      17.66     30.0

Carriers                            4,048    2,783    4,096     3,479

Carrier Bandwidth (KHz)       4.3125  4.3125   4.3125   8.625

Transmission Power(dBm)    +20.5   +14.5   +14.5   +14.5  **

Bandwidth (Mbps,bidirectional)   50       68       100       200

** DS  power

Ahhh, I assumed you were referring to IP profile rather than VDSL2 profile.

I think the profiles you have mentioned may be either the European or International versions.

I know that parts of America use the 8a profile, & I believe Japan also uses the 8c profile, possibly at full power.

I was on the UK's 8c profile until being switched to the 17a profile near the end of October.
My connection's instability & speed problems actually started at the end of July.

The 17a profile was implemented to enable BT's imminent speed doubling to 80Mb DS & 20Mb US.
I understand that due to the usually longer line lengths in the UK, VDSL2 power levels have been reduced to avoid cross-talk swamping of adjacent ADSL connections on a line-by-line basis. The longer the VDSL2 E-side, the lower the power needs to be near the cabinet.

My cabinet is probably around 4.5km from the exchange, so ADSL signal strengths will already be quite weak at that point, thus my cabinet's power level is probably quite a bit lower than at a cabinet say only 1 km from the exchange.

I must admit that I have wondered if maybe power levels were higher at first & possibly turned down following a month's worth of problem reports from ADSL users in the vicinity of my cabinet.
I have asked my ISP, but like everything else, they have no idea & BT won't tell them.

I still can't fathom out how I achieved very satisfactory & sustained speeds/stability at first, unless something is attenuating my signal more now than it was back in June/early July.
I can only think of reduced power as you have mentioned, or a high resistance issue.

For almost everyone else, the switch to the 17a profile brought a fairly substantial increase in DS speeds, at the expense of a slight lowering of US speeds.


Paul.

[burakkucat]

The longer the VDSL2 E-side, the lower the power needs to be near the cabinet.

sed 's/VDSL2/ADSL/'

I think you mean to say that "The longer the E-side cables, then the more attenuated the ADSL2+ signals will be. Thus the VDSL2 power needs to be adjusted to be comparable to the relatively weak ADSL2+ present at the PCP and, decreasing in strength, along the D-side cables on route to the customers' premises."

[Bald_Eagle1]

I think you mean to say that "The longer the E-side cables, then the more attenuated the ADSL2+ signals will be. Thus the VDSL2 power needs to be adjusted to be comparable to the relatively weak ADSL2+ present at the PCP and, decreasing in strength, along the D-side cables on route to the customers' premises."

Yeah, that too

[Bald_Eagle1]

As possible food for thought, don't SOME modems have an inbuilt facility to "ping" the DSLAM with some data, timing the return journey & possibly also estimating a line length?

Do you know if the HG612 has any such facility?

FWIW, I have just discovered that the Fritzbox 7390 does indeed report the VDSL Estimated Loop Length as shown in this TBB forum post:-
How to get a Fritz!Box 7390 working as a VDSL modem

i.e. VDSL Estimated Loop Length : 1120 ft

I know nothing about the Fritzbox 7390, other than a couple of users have it working as a VDSL2 modem for BT's 17a profile FTTC service & it can be accessed via Telnet (with a bit of a palaver).

Perhaps if we could somehow get hold of the code, we may be able to do the same via the HG612 (one way or another)?

Paul.

[asbokid]

As possible food for thought, don't SOME modems have an inbuilt facility to "ping" the DSLAM with some data, timing the return journey & possibly also estimating a line length?

Do you know if the HG612 has any such facility?

FWIW, I have just discovered that the Fritzbox 7390 does indeed report the VDSL Estimated Loop Length as shown in this TBB forum post:-
How to get a Fritz!Box 7390 working as a VDSL modem

i.e. VDSL Estimated Loop Length : 1120 ft

I know nothing about the Fritzbox 7390, other than a couple of users have it working as a VDSL2 modem for BT's 17a profile FTTC service & it can be accessed via Telnet (with a bit of a palaver).

Perhaps if we could somehow get hold of the code, we may be able to do the same via the HG612 (one way or another)?

Paul.

That's a very interesting find.  Some models of 2Wire* Motorola (e.g. 2210, 2241, 3347), models from the Siemens Speedstream series and the Ikanos-based devices from the Sagem f@st family (e.g. 3464) also have an algorithm in their firmware for loop length estimation. [1] [2]

The trouble is these are all closed-source proprietary algorithms and there is no formal mechanism within the xDSL specifications for calculating loop length.   When CPE does offer a loop length it must be an estimate derived from other line characteristics data, e.g. HLog(f).


cheers, a

[1] http://www.dslreports.com/forum/remark,22525726
[2] http://www.dslreports.com/forum/remark,22261547


* Thanks to Burakkucat for spotting the error.

[Bald_Eagle1]

Thanks for the link asbokid.

It also appears that the FRITZ!Box reports single attenuation values for DS & US rather than differing values for each band plan as per the HG612.

Paul.

[asbokid]

Thanks for the link asbokid.

It also appears that the FRITZ!Box reports single attenuation values for DS & US rather than differing values for each band plan as per the HG612.

Paul.

I'm sure that between us we could build a reliable algorithm for loop length estimation based on the line characteristics data provided by the CPE.

The maths behind Transmission Line Theory, the "Telegrapher's Equations", and the cable reference models initially looks daunting, but not overwhelmingly so.

cheers, a

[burakkucat]

Hmm.   

I wonder if the Fritz!Box is running a Linux kernel. If so, the code must be supplied, as per the GPL. 

Edit: That TBB thread, tells us that the OP was using a AVM Fritz!Box 7390.

AVM Computersysteme Vertriebs GmbH
Alt-Moabit 95
10559 Berlin, Germany
Phone +49-30-399-760
Fax +49-30-3997-6299
Email: info@avm.de

[Bald_Eagle1]

Hi b*cat,

Hmm.   

I wonder if the Fritz!Box is running a Linux kernel. If so, the code must be supplied, as per the GPL. 

Edit: That TBB thread, tells us that the OP was using a AVM Fritz!Box 7390.

Does this link provide anything of interest for those who actually understand these matters:-
Fritz!Box 7390 Linux kernel drivers

Paul.

[c6em]

In respect of BT plans......
I live on a private road and we the committee have the plans of all the underground services on the roads - including the BT one.  These were obtained quite some time ago (? 2000?) and at that time seemed quite freely available - well I had no problem in getting them but then I did have valid reason to require them.
The world has changed somewhat since then and maybe the plans of critical infrastructure items are no longer made so widely available.  The plans I have actually cover area larger than the immediate road. I keep the plans confidential though will supply extracts to residents re the services on their land if they are doing work on their property.

The plans show the main supply cable running down the verge of the road. The details are overdrawn onto the OS digimap series which itself was based on the old 1930's 25" to the mile large scale UK survey. The location of the joint boxes and pole are listed as are their ID/reference Nos.  Multipole overhead sections are also shown however the individual routes from the nearest pole to an actual property (ie the dropwire routes) are not shown, nor are the number of houses fed from a particular pole stated.

So it does show the route of the cable should you wish to follow it.

However the plan does not show the depth of the cable in the ground, or the exact position of the cable with reference to any fixed point, nor the cable type, or how many circuits are on each cable.   Some of the positions of the joint boxes are out horizontally by up to 5 meters but most are pretty accurate. As all the cable locally is armoured in ground not in ducts so I have no idea if ducted cables are shown differently.

In my area of rural home counties actually the cable run is pretty obvious and the map is of no real use - it's down the verge in all cases and then along undeneath a footpath alongside the road back to the exchange from the street box.

Hop this helps

 

[asbokid]

I just downloaded the 424MB tarball of tarballs containing the GPL'ed code for the Fritz!Box 7390 [1]

Unfortunately, there are no signs of a loop length estimation algorithm in the tarballs.

cheers, a

[1] http://download.avm.de/develper/opensrc/

[burakkucat]

Hmm.   

I wonder if the Fritz!Box is running a Linux kernel. If so, the code must be supplied, as per the GPL. 

Edit: That TBB thread, tells us that the OP was using a AVM Fritz!Box 7390.

Does this link provide anything of interest for those who actually understand these matters:-
Fritz!Box 7390 Linux kernel drivers

Apparently not, according to our Asbokid. (See above.) 

[Bald_Eagle1]

Unfortunately, there are no signs of a loop length estimation algorithm in the tarballs.

Ah well, at least it gave you something to do for a few minutes 

[asbokid]

Hi c6em!

Thank you for the info.  It's interesting that you obtained the data from BT.  That was very fortunate and useful to you. Access to the information is doubtless different when the land is your own.

The dog and I went on midnight manoeuvres the other night. We hoped to discover the route of the line from the house to the exchange. The idea was to trace the ducting by following the GPO/POT/Telecom/BT manhole covers. It was a modern version of the Hansel & Gretel story with an urban twist: we had to dodge the cops. They are on high alert for anyone acting suspiciously around manhole covers. Last month alone, the Great Drain Robbers nicked 200 covers.

Eight candidate cable routes were discovered, of which six are 'sane'. These routes were all paced out.  However, at each strategic road junction tracing the ducting any further was complete guesswork. Over three miles were trudged in the pursuit of science! Strangely, just one PCP was encountered enroute. A forlorn little thing it was, too.

cheers, a