Print

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

[GigabitEthernet]

I tried the same. They basically told me as long as line was within operating tolerance they won't raise a fault even though the banding was clearly unnesecary.

I got a reply which said the same. I've told them I'd like it removed again because the line doesn't need it - we'll see what they say.

I've researched this a little bit and the consensus is still not completely clear as to whether DLM ever removes banding itself. Some users on thinkbroadband seem to have had it removed automatically but I haven't seen any reports of that here.

There is something strange going on with my line though: the power on the upstream is often negative?

[burakkucat]

There is something strange going on with my line though: the power on the upstream is often negative?

A negative value is perfectly acceptable. Please remember that you are looking at units which display the power relative to 1 mW.

So positive values indicate powers greater than 1 mW whilst negative values indicate powers less than 1 mW. (And when 0 dBm is displayed, the power is exactly 1 mW.)

[GigabitEthernet]

But it never used to be negative. It's only since this banding incident that this has happened.

[burakkucat]

I can understand your logic in querying the power level but the setting is dynamic. The appropriate power levels are set by the two transceivers negotiating in the discovery phase. So there really is nothing to be concerned about. 

If it would help assume, as an experiment, the absolute power was initially 1.01 mW and now it is 0.99 mW. The difference between those two power levels is just 0.02 mW. Now calculate those two powers in terms of dBm and look at the difference.

[burakkucat]

Doing the calculations myself (as a result of a moment of doubt -- which seems to become more frequent with age) . . .

The equation is: n dBm = 10 x log₁₀(p mW / 1 mW)

When p = 1.01 mW, 10 x log₁₀(1.01 / 1) = +0.043 dBm (to three decimal places).

When p = 0.99 mW, 10 x log₁₀(0.99 / 1) = -0.044 dBm (to three decimal places).

Difference in absolute powers is 1.01 - 0.99 = 0.02

Difference in the relative powers (to 1 mW) is +0.043 - (-0.044) = 0.043 + 0.044 = 0.087

[GigabitEthernet]

Urgh

Hi
 
The predicted A Range for this line, thats under optimal conditions is 35 - 54mb
 
Your router is in sync at 49mb and this is within that range.  As such we or BT would not see this line at fault as it is behaving as predicted.
 
If unhappy with this, please follow our Complaint Process
 
Thanks
 
Karl.

[Chrysalis]

Presumably talktalk are been tight on engineer expenses.

I assume talktalk sell the service as "best efforts", that usually means lowest tier service, however it also should mean not artificially restricted.

I would send a polite reply stating that you simply want the service as is sold, basically for the line to run at the highest adapted sync rate, it is artificially capped excessively (12 db snrm with g.inp) and if they are not willing to follow due process of what is needed for a DLM reset then you want to leave your contract penalty free as they failing to fulfil those  basic conditions.

Of course the question is are you prepared to change isp to get this resolved? and I havent personally read the t&c's for talktalk, just made assumptions based on what I expect to be in there.

There is naturally a conflict in the way the service is sold, it is sold as a rate adaptive service, but also sold with estimated speed's, and its a question of which one overcomes the other, as the latter allows an isp to only consider a fault if the estimated speed is not achieved.

[GigabitEthernet]

We've got an exceptional deal with TalkTalk and so would not be willing to change.

The frustrating thing, is that my understanding is DLM won't even consider unbanding the line (if it ever does) until all counters are green, which the upload is not (it's yellow). G.INP would resolve this but DLM won't add it and so we're stuck in a vicious circle!

[burakkucat]

Just a few quick observations . . .

• An xDSL based service is provided as a "best effort".
• Perhaps WWWombat may be able to make use of the concept of the reduction in power as another empirical indicator of a circuit that is banded?

[ejs]

I assume talktalk sell the service as "best efforts", that usually means lowest tier service, however it also should mean not artificially restricted.

I would send a polite reply stating that you simply want the service as is sold, basically for the line to run at the highest adapted sync rate, it is artificially capped excessively (12 db snrm with g.inp) and if they are not willing to follow due process of what is needed for a DLM reset then you want to leave your contract penalty free as they failing to fulfil those  basic conditions.

I don't think that's what "best efforts" means. I think it's just wishful thinking that you've got a contractual right to the highest sync speed free of whatever things restricting it that you don't like.

I expect the terms and conditions will allow you to get the best speed that you line can achieve subject to all manner of things including the DLM.

[peterj]

I have to say a DLM system that never unbands a line that no longer needs intervention is broken and not fit for purpose.

[Black Sheep]

I have to say a DLM system that never unbands a line that no longer needs intervention is broken and not fit for purpose.

Agree ............. which is one like that doesn't exist.

[peterj]

I have certainly had enough consecutive green days to lift any other intervention - yet nothing. If it lifts at all (not convinced it does) then the settings are ludicrously conservative. It is pretty frustrating for anyone stuck on it.

[GigabitEthernet]

Agree ............. which is one like that doesn't exist.

Can you shed any light on the conditions for DLM to unband the line then?

[Black Sheep]

It's all been documented before on this forum  ...... maybe someone else can provide links to said info, I haven't got the time or wherewithal to go looking.