Print

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

[burakkucat]

In the broadest of terms, at the negotiation of sync phase, the DSLAM tells the modem how well it is "hearing" the modem's signal and the modem tells the DSLAM how well it is "hearing" the DSLAM's signal. One device does not control the other device's power output.

[kezzaman]

Lets just say, for some crazy reason i wanted to raise my routers output power a tad, how would one do that? 

[burakkucat]

I can "see" two options that would allow you to achieve your hypothetical desire:

(1) Modify your modem/router's hardware so that you have a physical "output power control knob" to turn.
(2) Modify your modem/router's firmware so that you can override the negotiated level in the software.

I don't know for sure if either of those suggestions could really be implemented but perhaps Asbokid would comment, please?

[asbokid]

I can "see" two options that would allow you to achieve your hypothetical desire:

(1) Modify your modem/router's hardware so that you have a physical "output power control knob" to turn.
(2) Modify your modem/router's firmware so that you can override the negotiated level in the software.

I don't know for sure if either of those suggestions could really be implemented but perhaps Asbokid would comment, please?

Presumably in theory it's possible to crank up the output power, i.e. the upstream power levels.   Except none of the DSL chipset makers release the source code for their DSP hardware drivers, nor even do they document their software interfaces.  For those reasons, there's not much chance of succeeding.  Success would also only relate to the upstream power.   The downstream performance of an xDSL connection is determined by the performance of the analog front end (AFE).   The AFE is the circuitry that recovers and digitises the symbols from the QAM-modulated analog signal on the line. However the AFE will be already be running optimally.

cheers, a

[kitz]

I think theres some confusion going on here (or at least Im confused!).
I'm going to recap a few items to see if Ive got this straight for what you are trying to do.


- It would appear that your router doesnt display the downstream output power - some dont. 
- The upstream and downstream output power isnt the same.  A very rough guide is upstream will be about 1/2 the downstream on adsl2+ but this doesnt always follow.
- Changing your DSLAM line profile doesnt affect the output power.
- Requesting a change in your Target SNRm shouldnt make any difference to output power.
- Line length can affect your output power but it all depends if your line is syncing to its maximum capability.



---------

Why would you be wanting to change the output power?

The downstream is controlled at the MSAN.  Normally its around 18/19dBm but the MSAN can  'tweak it up a bit' > 20dBm if the line needs it..  or the MSAN will reduce the power if it senses that the line doesnt need it.  ie a short line syncing at 8Mbps.


------------------

I have an issue with my connection which sometimes goes away, most recently was about 2 weeks ago when it went away and liturally the only thing i could see different with my stats was that the upstream output power was 0.5 higher at 9.5, since then its been at 9 and hasnt budged. Do u think an increase of 0.5 could have an affect on someones connection?

Not really in this case - it could have been a temp blip and output power can vary slightly  -  also Remember upstream and downstream are independent. 
Many routers also only report rounding to 0.5dB..  its not impossible that tiny fluctuations and roundings can make a difference with 0.5dB.

Is it possible that i could be experiencing gaming online issues because my upstream is suffering from crosstalk at the exchange?

doubtful.. I refer you back to this post which has links to more info about SNR/SNRM and the suggestion of using routerstats to see what is going on for your line.

I think you need to be looking at whats happening with your SNRm before thinking about output power., as any fluctuations in SNRm is an indication of noise on the line, which is what could be causing problems.   If the SNRm is too low then this could very well cause problems with gaming etc


----
edited to add..  asbokid posted whilst I was posting.

[kezzaman]

Ive been through alot with my connection and after all the things ive tried to sort it out with no luck, im starting to think it might be an issue with the upstream, so im not concerntrating to much on my downstream at the moment.

Ive just began running routerstats lite on my laptop, i had to use lite because the full version didnt work with my router dsl-2780.

[kezzaman]

Hello,
 
Heres my routerstats on the downstream, i couldnt get upstream. Can you tell me if they look normal please?

[kitz]

>> i couldnt get upstream.

Thats a shame if you say its your upstream that is concerning you. 

>> Can you tell me if they look normal please

That graph is only over a short period of time ( 1 hour) but it seems to be a fair bit of constant fluctuation between 10-12dB. 
Whilst indicative that there is some sort of noise going on, most lines should be able to cope with 2dB fluctuations..  that graph over the short period doesnt show anything why your target SNRm would be at 12dB.

It probably needs monitoring over a longer period to see whats really going on..  some lines can be more problematic in the evenings when more people are at home and thats when any effects of EMI are most often seen.

[kezzaman]

Thanks Kitz,

My snr doesnt need to be at 12db i was just trying something out, normally im on a 9db profile. I noticed that it never went above 12db only downwards, do u think thats just because it doesnt need to be that high?
I do have a netgear which will work with the full routerstats program, ill get it setup in the week, means going in the loft   
and then ill get more results over a longer period of time.

[Black Sheep]

Regarding 'Output power'. This really shouldn't have any impact on you as EU's, but 'Low Power Mode' (L2) is now being applied at the MSAN's.

This has been pretty quick to come to fruition, and has taken 18 months from concept, through trials, to implementation, and what happens is when there is little (under 128Kbps) no 'traffic flow' seen by the MSAN for 127 seconds. It basically goes to sleep.

A result of this, we are experiencing as engineers, is that our testers are not drawing heavy traffic when plugged into the circuit, and as such we are actually seeing the drop from nominal synch speed, down to figures between 100-200Kbps. The first time this happened to me, my eyes nearly popped out of my head !! We obviously, had not been briefed about this new change. A couple of guys have spent hours chasing a 'perceived' fault condition, when all the time it was just L2 Mode switching in.

The minute the EU accessess the laptop/PC again, the nominal synch is returned instantaneously so you shouldn't in theory notice anything.

This has only been 'in play' for a couple of weeks, and already a main bug bear we have found is with the 2-wire BT Bus Routers. The issue is they tend to lose their PPP Session when it goes into L2 mode, which is damn frustrating for the EU. The only fix at present that I am aware of, is to install the latest BT Business Hub V3.

I believe the guys at Martlesham are still tweaking this ??

[Bald_Eagle1]

Hi BS,

Is this only relevant to ADSL connections?

I do notice from my graphs that DS Output Power fluctuates a lot more on my connection than any others I have seen.

I have taken that to be that power is increased whenever my connection is "struggling" to maintain SNR levels.
These fluctuations can be seen minute by minute, sometimes by almost 1 dBmV.

The 2nd graph is from a much longer FTTC connection (1600m or so), only able to achieve 18Mb sync speed, yet Output Power hardly fluctuates at all, even over a 4 day period

[c6em]

Interesting:
I wonder what the line monitoring programs such as Routerstats/Routerstatslite and the DMT tool will make of the gyrations in sync speed.

The SNR margin may start to show interesting behaviour as well.
If you are using the web while all around you are not the reduced power should manifest itself as a greater SNR margin.
On the other hand the action in raising/lowering power on other lines will cause increased cable cross talk on yours - hence reducing SNR margins while its happening.
Whether you will see this in practice is another matter....

[Black Sheep]

BE (apologies first re: the other thing I was attempting to do for you ....... have been extremley busy of late)

Regarding the 'Cool Broadband' issue. I know for fact in relation to ADSL, it can only be applied to 21CN circuits ..... NOT 20CN. I don't know about VDSL though pal ??

As I say, this thing was thrust upon us with no communique whatsoever !! When I know more, I'll let you know.

[c6em]

Here is a forum link I've found to a BT overview of the 'swap to low power' concept, plus a pdf link to a more technical document describing how it works and the issues surrounding it.

http://www.thinkbroadband.com/news/5150-bt-wholesale-adsl2-to-go-all-cool.html

[kitz]

>> This really shouldn't have any impact on you as EU's, but 'Low Power Mode' (L2) is now being applied at the MSAN's.

Thanks BS for that info.  I'd previously seen that BT were meant to be bringing in 'Cool Broadband'.  They do indeed seem to have rolled it out PDQ

>> I have taken that to be that power is increased whenever my connection is "struggling" to maintain SNR levels.

Lots of bit swapping will also cause changes in output power.