Print

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

[konrado5]

What do you think about results?

Code: [Select]

Test Time (sec):    360
Total Transferred Bits:    0x00000001404A8D80
Total Error Bits:    0x0000000000000490
Error Ratio:    2.17e-07

[konrado5]

Code: [Select]

Test Time (sec):    360
Total Transferred Bits:    0x00000001404A8D80
Total Error Bits:    0x00000000000001C4
Error Ratio:    8.41e-08

[burakkucat]

I think they are incredibly good results, showing that you have a very good line. Considering that you have set the target SNRM to what is, essentially, an unworkable value for a UK line, those error ratios are insignificant.

To be honest, I have no idea why you persist in keeping the target SNRM set to such a low value other than, perhaps, because you can. 

My advice is to set the target SNRM to 3 dB and then look at the line statistics. They will still be far better than what I can achieve. 

[konrado5]

I think they are incredibly good results, showing that you have a very good line.

My hlog shows my line is poor.

To be honest, I have no idea why you persist in keeping the target SNRM set to such a low value other than, perhaps, because you can.

If I can higher synchronization rate keeping stability why I'd lower it?

[kitz]

I think they are incredibly good results, showing that you have a very good line.

My hlog shows my line is poor.

That hlog is NOT poor at all..  its quite good.   
But Im sure weve been over this before...  you have a bit of background mush at 410-475 but that is hardly unusual.   


Based on your latest findings about a possible DSLAM filter, then it probably proves that you ISP is right to filter these tones.  Whilst it may not be bothering your line too much, there is a strong chance that many hundreds of other users on your DSLAM a bit nearer to the source of the RFI could be suffering quite badly without the filter.

Does the black cat concur with my thoughts in the last paragraph?

[konrado5]

kitz: burakkucat noticed "lazy" roll in the Hlog graph and:

Looking at your Hlog graph, it does seem to be somewhat abnormal. 

It does not show a smooth decay versus frequency that we would normally expect to see. The undulations may be the result of impedance mismatch, where different sections of cable (of different characteristics) have been joined together.

It is curious. Poor Hlog graph and very stable circuit.

Based on your latest findings about a possible DSLAM filter, then it probably proves that you ISP is right to filter these tones.  Whilst it may not be bothering your line too much, there is a strong chance that many hundreds of other users on your DSLAM a bit nearer to the source of the RFI could be suffering quite badly without the filter.

I'm sure nobody in my town broadcast on 160m band. Moreover, before September 2013 I've had probably all used tones.

[kitz]

OK now Im getting tired of this.    I have REPEATEDLY said that those tones to me show some sort of background noise from RFI/EMI.

To me it displays signs of some sort of interference.  Now I refuse to discuss that again, because over the past few months, several people have advised you what those possible causes could be. You have dismissed every possible cause - and there have been numerous suggestions made. You dismissed local radio hams, but how do you know there isnt one for sure..

but Im afraid that there is nothing that your ISP will be able to do about them if it is radio mush. 

Due to the fact that you say sometimes a very low snr my gut instinct would say its more likely a local interference...  Or just plain mush that happens for no apparent reason on many many lines and is considered within the realms of adsl normality.

What I think you are missing the point on is that because of the way adsl works it is pretty normal for most lines to see some tones that are affected by something or other.   In fact its far more common to see some tones missing or dipped, its rare that there is a straight line decline without some sort of interference or other.  It only usually happens on the very short and good lines.  Even just normal background noise mush causes peaks and troughs.

I would say that is far more likely to be local interference of some sort..  ie the background mush talked about in the past..

Stop. it. right. there.!   

That has already been explained.. three times..  its background mush.

Looking at the graphs, hlog..confirms my previous post & as B*cat says - looks like background mush.

All this before we knew that poland does something we dont see in the UK and thats use Power Cutback to filter and protect against tones used by radio hams.

One of the guys on here (sorry cant recall who it was now) found a document that stated those tones were in use by radio hams in your country.   Then you yourself found a whitepaper that showed how power cutback could be used to protect against radio hams.

You state you arent aware of radio hams broadcasting, but you can never be sure as its something any radio enthusiast can do from within their own home.  A give away may be a large aerial in their back yard, but it may just as likely be hidden from public view.


I think its a pretty sure fire bet that someone is broadcasting on those frequencies either in or near your town.  Those squiggles on your hlog prove that there is some RFI there but you only get it slightly.  HOWEVER people nearer the location will get a stronger and much more interference causing their line to consistently drop.  It wouldn't surprise me in the slightest if your ISP (or the dslam manufacturer) has put that block in place to protect the lines that were having problems with the RFI.

Its tough luck that the filter affects your line. but there isnt going to be anything you can do about it.   The ISP are going to balance stability of 100s of lines against one line which looses a bit of sync speed.  Ive repeatedly told you that here in the UK 10s of 1000's of BE users had to live with the loss of available sync speed because the  ISP disabled (not filtered) a block of tones on ALL DSLAMs for the maritime tones regardless if individual users were affected or not. 

It happens. Live with it and be happy that you actually have a decent line.   

Im done

[konrado5]

1. my QLN is almost flat on these tones.
2. burakkucat noticed something not related to these tones: impedance mismatch.
3. By comparison, someone else in Poland has similar HAM filter but he has much better Hlog (despire he has phone line in air, I have line under the ground).
4. Where are lines in UK placed? Under the ground or in air?

Best regards
konrado5

[kitz]

1. my QLN is almost flat on these tones.

So?   
That just confirms that it is power cutback that has been used to filter the tones against RFI.

Look at my graphs in this post  See how my QLN is flat in the U1 band because BT has applied power profiles for short lines to protect against crosstalk on VDSL on the whole U1 band.

Although (afaik) we dont have any ISPs that use power cutback on specific RF tones in the UK like you do in Poland to protect against HAMs.  BT still uses powercut back for line profiling to prevent crosstalk.  This affects a whole band, not just specific tones.

  burakkucat noticed something not related to these tones: impedance mismatch.

No.  He offered the suggestion  "The undulations may be the result of impedance mismatch". 
B*cat has gone above and beyond the call of duty to try and offer suggestions why you may see those squiggles on some lines.  Im pretty certain that he has offered RFI and background mush, and that suggestion was only made when you denied it could be RFI.

By comparison, someone else in Poland has similar HAM filter but he has much better Hlog (despire he has phone line in air, I have line under the ground).

So he may live further away from the radio ham and thereby not affected as much.  The radio ham in his town may not be broadcasting as strong.  RFI will never affect 2 lines in exactly the same way.   

. Where are lines in UK placed? Under the ground or in air?

Both.  Depends where you live and how old the property is. 

 

[konrado5]

kitz: undulation in my Hlog are on much lower frequiences than HAM band. For example there is "roll" on the beginning of download band.

That just confirms that it is power cutback that has been used to filter the tones against RFI.

And also confirms lack of RFI.
In fact there are amateur radio enthusiasts in my town, but nobody broadcast at 160m band.
I've found only 142.312.5 Mhz amateur radio.
http://aprs.fi/info/?call=SP9KLF
I've talked also to amateur radio enthusiast and he confirmed: nobody broadcast at 160m band. Some people broadcast at 2m band.

EDIT: Finally, my ISP confirmed that they set PSD masks for HAM band (according to hardware vendor reccomendation).

Best regards
konrado5

[burakkucat]

I am getting thoroughly confused . . . 

annoyed . . . 

and sad . . . 

An Hlog graph is a plot of attenuation versus frequency. In general, one can see indications of abnormalities in the physical properties of a circuit by viewing its Hlog curve.

When first shown konrado5's Hlog graph, my initial thought was that it is somewhat abnormal. I still hold that view -- it is somewhat abnormal. Why do I make that statement? Because of the "slow lazy roll", the "undulations" that are present in the curve. What could possibly be a cause of the varying attenuation over the frequency range in question? Changes in impedance (i.e. the AC resistance) along the metallic pathway, between the CPE and the DSLAM, would be a good guess. What could possibly be the cause of such impedance changes? One possible answer is that the metallic pathway of the circuit in question could be made up of various lengths of different poundage cables.

Everything I have typed in the above paragraph is due considered speculation that fits in with the observed result. Other than travelling to Poland and:

(1) physically examining the entire length of the circuit
(2) connecting various items of test equipment to the circuit & observing the results obtained
(3) discussing the rules/regulations that govern the implementation of xDSL, in Poland, with the relevant statutory authority

then there is nothing more that I can say. 

No more questions, please. 

My advice is to set the target SNRM to 3 dB (which is the minimum sensible overhead) and just make use of the broadband service that is provided.

[kitz]

Hang on which graph are we looking at.

I was looking at this one

It looks pretty damn near perfect to me aside from what could be a small amount of EMI/RFI/mush


TBH b*cat Im sick to the back teeth of it all.   Ive been bombarded with so many figures, questions about trivial changes that dont matter... when overall according to this hlog it looks good.

It just goes on and on and on and on...  and it makes my head spin.

I feel I have nothing else to offer to this conversation.  The whole escapade has exhausted me.   

[konrado5]

Above graph is not my graph. Following there is my graph. How is it possible: poor Hlog graph and very stable circuit on SNR margin ~1 dB? Phone line is under the ground.

[burakkucat]

I've had a quick search to find a reference to konrado5's Hlog graph. It is the first graph in this forum post and is also attached, below.

[konrado5]

burakkucat:  How is it possible: poor Hlog graph and very stable circuit on SNR margin ~1 dB? Phone line is under the ground.