Print

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

[seaeagle]

I have recently changed from ADSL2 to FTTN with a hop of approximately 150m to our premise. I have signed up for a 12/1 D/U Mb/s as suitable for our requirement.

Using a Netcomm NF18acv router which I can Telnet into to get ADSL data out.  This way I get the data out and use a bash script and gnuplot to visualise the result on my Linux desktop.

The result is not what I expected but may be true.  I have no problem with getting my alloted speed it is more curiosity that has sent me searching for practical data/graphs in this area.  A number of years ago I rembered reading Kitz information about ADSL and also saw my first Line connection analyser, think it was a German produced program.

So I spent many hours fumbling with bash programming until I got to a stage that I can almost understand what I am doing.

The SNR has me baffled and my first concern is the range of tones 2700-3200.  What is going on?

Oooops, looks like I can not attach a picture.  I can browse to a picture but I see no upload.  I will post this while I try to find out what I have not read about being new to forum   

Added
Hmm appears that it was only the preview that did not show picture.

[kitz]

Hi and welcome

>> think it was a German produced program

Possibly DMTtool?

>> The SNR has me baffled and my first concern is the range of tones 2700-3200.  What is going on?

It looks like ~1950 to ~2800 is one of your upstream bands.   

Your Hlog looks good indicating no problem with the physical line.   
Your QLN also looks good and I cant see anything in there which would indicate noise

Not quite sure whats going on ~2800 to 3150
Your SNR shows possible RFI/EMI in those last tones... but what is unusual is that I would expect your QLN to mirror  any noise on those tones. 

QLN (& hlog) are a one shot graphs taken during the initialisation stages of sync.  So it could be the noise only started after sync.   Its not actually affecting your bit loading and your modem is estimating fairly decent speeds.   

Don't know how many others are on your node... but I would expect perhaps a slightly higher max rate at 150m.    I cant see obvious signs of crosstalk from others on the same node.  Perhaps a tiny bit in the adsl range at but nothing that stands out.  The other thing to bear in mind is that because you are only on 12/1 you are not using the line anywhere near to its full potential so I would expect there to be a fair bit of power cut back and some spectral shaping going on. In cases such as this it is possible that if you were ever to upgrade to say 80/20 then you would be given a bit more power... which would in turn ramp up your SNR and allowing for slightly better bit loading.

[seaeagle]

Thank you for replying. Yes it was early version of DMTool. What I have read in your answer and re-reading the page:

http://www.kitz.co.uk/adsl/adsl_technology.htm

Reading above I have trouble understanding that a SNR value for a Tone can show up as Zero if all the tones are evaluated, there seems to be no reason. From ~2550 and onward range, the QLN seem seem quite  'smooth' and yet there is SNR is 0. (you have pointed that out)

Since the the QLN is nice and smooth right through the 'hash' range as well this is interesting.  I will make my program portable and see if I can get readings from another FibreNode. I do have a chart from another node not that far away, the graph made with another home made program.

I am in a quiet suburb almost rural, located in Tasmania south of mainland Australia, and forgot the big Dairy factory a few hundred meter up the road. Though it is my understanding it is on its own mains high voltage tap so our mains in the neighbourhood should not be affected.