Kitz Forum
Broadband Related => Broadband Technology => Topic started by: Weaver on August 14, 2015, 08:52:51 PM
-
Maximum reportable attenuation = 63.5 dB, was what I read somewhere, in that modems report this figure when the attenuation is actually much worse.
I wonder what my line plus modem really scores? How bad is the true figure?
It's about four maybe nearer five miles long, presumably thickish copper with no problems in it as the performance figures wouldn't be as good as they are.
--
PS: Correction - I originally used a minus sign for attenuation dB, god knows why, corrected it to be positive. I changed the term infinity to maximum
-
Actually, the probable figure might be close to 63.5 anyway, as the performance is relatively good, and recalling the days when I had a Netgear DG834v1 modem/router, it used to read 60.5 or 61dB ds attn. The figure has probably got worse since then because I now use higher tones more.
-
I'm not sure where you've got the minus sign from, but the 63.5 dB value is the maximum downstream attenuation value that can be stored in certain ADSL1 messages. Many modems will display 63.5 as the maximum when operating on ADSL1, probably because they store it in a data structure that corresponds to the ADSL1 messages.
The AR7 DG834s can be a little different to other modems, the attenuation mine displays varies with the speed, connected at a higher speed, it displays a higher attenuation. It's probably displaying the attenuation based on the tones in use.
-
> attenuation based on the tones in use
Indeed, that's what I assumed. Frequency-dependent, averaged, is the meaningful way to do it.
As for the minus sign, no, I just had too many pain killers again. Let's make it positive. Original power over far end power rather than the other way round.
-
@ejs - you wouldn't happen to have some reading matter for me about that particular field in the protocol?
-
Only the ITU-T G.992.1 (http://www.itu.int/ITU-T/recommendations/rec.aspx?rec=4718) Technical Recommendation itself, R-MSG-RA in particular. No, I don't understand most of that document, but the ITU documents can be a useful reference.
Of course a modem could display or store the attenuation of the received signal before encoding it into such a limited number of bits, but most don't.
-
I can't see the relevance to Infinity.
-
> I can't see the relevance to Infinity.
"Potentially unlimited", or something like that, would have been a much better expression.
So the only bit of the title that was correct was the number. :-)
-
Ahhh.. gotcha - I too was confused with the term 'Infinity' - thinking huh but adsl and vdsl attenuation wont be the same. You mean maximum reported 63.5dB attenuation (http://www.kitz.co.uk/adsl/linestats.htm#attenuation_63dB).
ejs has already pointed out ITU-T G.992.1 which defines the test parameters for DSL.
Some modem manufacturers appear to have taken it literally as to meaning they only have to show up to 63.5dB and therefore stopped dead at this figure.
Not all manufacturers took it quite so literally, and there are some modems which will display figures over and above this.
-
I wonder what the record is then, the worst ever true reported figure?
-
I remember seeing 71dB once on a 2Wire 2700HGV when the attenuation at 300KHz was reported.
-
Certain modems will indeed display figures higher than this. Before FTTC was activated on my line it was showing attenuation of 72dB. Needless to say I wasn't streaming anything in HD on that connection.
-
@lcl00 - Wow! perhaps that's where I am really.
What speeds were you getting?
-
Btw is there a reasonably accurate table or function to link length with attenuation?
-
We used to say that the UK average was about 14 dB per kilometre for ADSL1, but it can vary quite a bit, depending on the type and size of the cable used, and also other issues such as the number and quality of connections.
-
That's what I read somewhere. Since I am more than 6 km from the DSLAM, that makes my ds attn an incredible 84 dB or even worse. Can't be possible, can it?
-
Perhaps BT originally put in superb thick copper for such long lines, so that voice would be ok?
Does that ever happen?
-
That's quite possible. I believe there was a mention of this on the forum fairly recently.
-
BT will use thicker cable for properties a long way from the cabinet. About 800m of my D-side is 0.9mm copper and the last 500m or so is 0.6mm (I tried to push for 0.9mm, but they said it couldn't be run overhead). However, the E-side was likely aluminium, which is why the figure was so woeful. I think I read somewhere that BT tried to ensure that all of their network was no longer than the equivalent of 6.5km @ 0.5mm, but I could be pulling that out of my rear.
14dB/km must be for thinner and possibly more degraded lengths of copper, or maybe a very conservative average. On my old routing I was 5.9-6.0km from the exchange and my attenuation on all 3 lines was between 56-57dB on ADSL1, which wouldn't correspond very well to those measurements. Realistically once it hits that maximum of 63.5dB, it probably would be pretty difficult to estimate what the attenuation figure actually is.
-
The only real numbers I have are from back in the days when I was probably not using the higher bins on fixed 500k ADSL. And then I scored something like 61 dB ds attn. and nowadays, not quite reaching the 2272 ds sync rate threshold for 2 Mbps, it's gone off the end of the scale possibly only slightly for all I know. I have often hit 2272 but only have only held it for a few days to a week.